(19) TZZ¥¥ T · a camera system or by ultrasonic sensors. The detected movement can then be used to re-compute the sensor locations by compensating the movement accordingly. In

Printed by Jouve, 75001 PARIS (FR)

(19)E

P3

354

197

A1

TEPZZ¥¥54_97A_T(11) EP 3 354 197 A1

(12) EUROPEAN PATENT APPLICATION

(43) Date of publication: 01.08.2018 Bulletin 2018/31

(21) Application number: 17153752.5

(22) Date of filing: 30.01.2017

(51) Int Cl.:A61B 5/044 (2006.01) A61B 5/046 (2006.01)

(84) Designated Contracting States: AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TRDesignated Extension States: BA MEDesignated Validation States: MA MD

(71) Applicants: • Karlsruher Institut für Technologie

76131 Karlsruhe (DE)• Städtisches Klinikum Karsruhe gGmbH

76131 Karlsruhe (DE)

(72) Inventors: • Oesterlein, Tobias

76227 Karlsruhe (DE)• Dössel, Olaf

76646 Bruchsal (DE)• Schmitt, Claus

69121 Heidelberg (DE)• Luik, Armin

76187 Karlsruhe (DE)

(74) Representative: Bittner, Peter et alPeter Bittner und Partner Seegarten 2469190 Walldorf (DE)

(54) METHOD AND SYSTEM FOR IDENTIFYING POTENTIAL ATRIAL FLUTTER AREAS FOR MEDICAL DECISION SUPPORT

(57) A decision support computer system, compu-ter-implemented method and computer program productfor supporting diagnostic analysis of potential ablationareas to cure atrial flutter. An interface component (110)receives an electrical reference signal (RS) generatedby a reference sensor (CSCS). The reference signal (RS)provides a time reference for electrical activity of one ormore atria of a patient wherein the time interval betweentwo subsequent electrical activities is referred to as basiccycle length (BCL). Further, it receives a plurality of elec-trical signals (P1 to Pi) generated by one or more furthersensors (S1 to Sn) wherein the plurality of electrical sig-nals (P1 to Pi) relate to a plurality of locations in the oneor more atria. The plurality of locations includes locationsdifferent from the location of the reference sensor(CSCS). A signal analyzer component (120) determines,for each signal of at least a subset of the received signalsoriginating from locations within a selected area of theone or more atria, when electrical activity occurs. Further,it determines, for a plurality of time points within the basiccycle length, size values for respective areas of activetissue by aggregating the size of surface elements asso-ciated with the locations where the respective signalsshow activity at the respective time points. A visualizercomponent (130) sets a visual property value for at leasta subset of the surface elements wherein the subset in-cludes surface elements associated with locations defin-ing a particular area of active tissue which is smaller thanactive tissue areas associated with other locations. The

visual property value indicates, in a visualization of theselected area, the particular area of active tissue as apotential ablation area to cure atrial flutter.

EP 3 354 197 A1

2

5

10

15

20

25

30

35

40

45

50

55

Description

Technical Field

[0001] The present invention generally relates to elec-tronic data processing, and more particularly, relates tomethods, computer program products and systems foridentifying potential atrial flutter maintaining areas basedon clinical sensor data.

Background

[0002] Atrial flutter can frequently be observed as ei-ther index arrhythmia or as organized tachycardia follow-ing the ablation of atrial fibrillation (AFib). In the lattercase, atrial flutter (AFlut) often is resistant to pharmaco-logical treatment and cardioversion, but as well seen asintermediate step before normal sinus rhythm (SR) isachieved. In both cases, however, numerous studieshave already demonstrated consistent results with re-spect to the underlying mechanisms and their anatomicallocations. This understanding and the resulting treatmentdecisions have led to success rates in the order of 80 to100% when catheter ablation is applied to terminateAFlut. The understanding of the patient-specific variant,however, is the central prerequisite of successful abla-tion. Details regarding the medical foundations can befound in the following references: "A classification of atrialflutter and regular atrial tachycardia according to electro-physiological mechanisms and anatomical bases; aStatement from a Joint Expert Group from The WorkingGroup of Arrhythmias of the European Society of Cardi-ology and the North American Society of Pacing andElectrophysiology (by Saoudi, N.; Cosio, F.; Waldo, A.;Chen, S. A.; lesaka, Y.; Lesh, M.; Saksena, S.; Salerno,J.; Schoels, W.; European Heart Journal; 2001; 22;1162-1182)"; "Atrial tachycardia after ablation of persist-ent atrial fibrillation: identification of the critical isthmuswith a combination of multielectrode activation mappingand targeted entrainment mapping; (by Patel, A. M.;d’Avila, A.; Neuzil, P.; Kim, S. J.; Mela, T.; Singh, J. P.;Ruskin, J. N.; Reddy, V. Y.; 2008; Circulation. Arrhythmiaand Electrophysiology; 1: 14-22)"; and "Characterization,Mapping and Ablation of Complex Atrial Tachycardia: In-itial Experience with a Novel Method of Ultra High-Den-sity 3D Mapping; (by Schaeffer, B.; Hoffmann, B. A.; Mey-er, C.; Akbulak, R. O.; Moser, J.; Jularic, M.; Eickholt, C.;Nuhrich, J. M.; Kuklik, P.; Willems, S.; 2016; Journal ofCardiovascular Electrophysiology; 27: 1139-1150)".[0003] Currently, this diagnostic process is primarilybased on the experience of the physician. Although com-prehensive electrogram data can be acquired from thecomplete atrium using an electro-anatomical mappingsystem (EAMS) with multi-polar mapping catheters, onlytwo automatic analysis techniques are known. First, themaximum bipolar voltage of electrograms (EGMs) isevaluated within one cycle. Regions which exhibit voltagevalues below a certain threshold are considered scar,

potentially being part of an anatomical obstacle. Typicalthreshold values for atrial myocardium are in the rangeof 0.5mV for zones of low voltage and 0.05mV for densescar. Second, the local activation time (LAT) is annotatedin each signal. After all signals are synchronized, the LATreflects the global excitation pattern. This static imageallows to gain a first impression of the course of atrialactivation.[0004] However, a number of additional features areassessed by physicians in the diagnostic process andtypically evaluated manually by visual inspection. Con-sidering the analysis of individual signals, these comprisethe presence of double potentials, and evaluation of frac-tionation and duration of activity in each signal. From amultichannel point of view, additional features can be de-termined, like the mid-diastolic activity, area-based cyclelength coverage and the critical isthmus. In addition tothe references above by Patel et al. and Schaeffler et al.,further medical details regarding the additional featuresare disclosed in the references: "A deductive mappingstrategy for atrial tachycardia following atrial fibrillationablation: importance of localized reentry; (by Jais, P.;Matsuo, S.; Knecht, S.; Weerasooriya, R.; Hocini, M.;Sacher, F.; Wright, M.; Nault, I.; Lellouche, N.; Klein, G.;Clementy, J.; Haissaguerre, M.; 2009; Journal of Cardi-ovascular Electrophysiology; 20; 480-491)", "Treatmentof macro-re-entrant atrial tachycardia based on electro-anatomic mapping: identification and ablation of the mid-diastolic isthmus; (by De Ponti, R.; Verlato, R.; Bertaglia,E.; Del Greco, M.; Fusco, A.; Bottoni, N.; Drago, F.; Sci-arra, L.; Ometto, R.; Mantovan, R.; Salerno-Uriarte, J.A.; 2007; Europace : European Pacing, Arrhythmias, andCardiac Electrophysiology : Journal of the WorkingGroups on Cardiac Pacing, Arrhythmias, and CardiacCellular Electrophysiology of the European Society ofCardiology; 9; 449-457)", and "Selection of Critical Isth-mus in Scar-Related Atrial Tachycardia Using a New Au-tomated Ultrahigh Resolution Mapping System; (by D.G. Latcu, S. - S. Bun, F. Viera, T. Delassi, M. El Jamili,A. Al Amoura, N. Saoudi; 2017; Circulation. Arrhythmiaand Electrophysiology; 10; online ahead of print".[0005] The flood of clinical data measured by corre-sponding sensors is overwhelming for any physician anddoes not appropriately support a physician in his/her ef-fort in diagnosing atrial flutter types and locations for apatient.

Summary

[0006] Therefore, there is a need to provide a diagnos-tic decision support solution with a corresponding methodto assess these parameters automatically in order to sup-port a medically trained human (e.g., a physician) in un-derstanding the flutter circuit for improved diagnostic de-cision making as a basis for successful ablation.[0007] This technical problem is solved by the featuresof a computer system, a computer-implemented methodand a computer program product as disclosed in the in-

1 2

EP 3 354 197 A1

3

5

10

15

20

25

30

35

40

45

50

55

dependent claims.[0008] In one embodiment, a decision support compu-ter system for supporting diagnostic analysis of potentialablation areas to cure atrial flutter includes an interfacecomponent to receive an electrical reference signal gen-erated by a reference sensor. Typically, a coronary sinuscatheter sensor (CSCS) is used as a reference sensor.However, a reference sensor may also be placed at alocation different from the coronary sinus in the patient’satria. In one embodiment, an ECG sensor may also beused as a reference sensor. The reference signal pro-vides a time reference for electrical activity of one or moreatria of a patient wherein the time interval between twosubsequent electrical activities is referred to as basic cy-cle length (BCL). Time reference, as used herein, meansthat any delay times of signals measured by the furthersensors at particular locations are measured in relationto the signal measured by the reference sensor. The in-terface further receives a plurality of electrical signalsgenerated by one or more further sensors wherein theplurality of electrical signals relate to a plurality of loca-tions in the one or more atria wherein the locations in-clude locations different from the location of the referencesensor.[0009] In one embodiment, the reference sensor mayprovide a position reference. Position reference, as usedherein, means that the position of each further sensor isknown relative to the position of the reference sensor.Often, the CSCS is used as the position reference. Typ-ically, the locations of a few sensors in the heart are de-termined via magnet coils mounted at the tip of the cath-eter and underneath the patient whereas the other sen-sors are located using the principle of potential divider.Once the position of the reference sensor is determinedthe positions of other sensors are known relative to theposition of the reference sensor. If the patient moves, therelative positions can be determined because the move-ment of the reference sensor is detected. It is also pos-sible to determine the movement of the patient by othermeans. For example, the movement may be detected bya camera system or by ultrasonic sensors. The detectedmovement can then be used to re-compute the sensorlocations by compensating the movement accordingly.In these implementations, the position reference functionof the reference sensor is optional. Further, the interfacereceives a plurality of electrical signals generated by oneor more further sensors wherein the plurality of electricalsignals relate to a plurality of locations in the one or moreatria. The plurality of locations is different from the loca-tion of the reference sensor (CSCS). In other words, theelectrical signals received by the one or more further sen-sors illustrate how the excitation wave propagatesthrough the atrium of a patient before and after passingthe coronary sinus as detected by the reference sensor.The signals may be received by the interface directly fromthe sensors, thus enabling a live monitoring of the atriaactivity or it may be retrieved from a database whichstores historic measurement data of the sensors.

[0010] Further, a signal analyzer component of the sys-tem determines, for each signal of at least a subset A ofthe received signals originating from locations within aselected area of the one or more atria, when electricalactivity occurs. Typically, in case of atrial flutter, an elec-trical excitation wave is propagating through, and thusdepolarizing, the tissue of the atrium. The surface of theatrium which is depolarizing at a particular point in timeis referred to as the area of active tissue at this point intime. Dependent on the size of this area a plurality of thereceived signals corresponding to respective locationsin the patient’s atrium can show activity at the same time(simultaneous activity). The atrium is typically visualizedfor the medically trained person through a 3D graphicalobject. Typically, a 3D object surface is generated usinga plurality of surface elements (e.g., polygons such astriangles or other shapes). For optimized processing, themesh surface elements can be converted between dif-ferent forms (triangles, tetrahedrons, etc.) or increasedor decreased in density by increasing or reducing thenumber of surface elements. Each surface element canbe assigned to a respective sensor location. For exam-ple, the respective sensor can be defined as the closestsensor in space. The signal analyzer then determines,for a plurality of time points within the basic cycle length,size values for respective areas of active tissue by as-sessing the aggregate size of the surface elements as-sociated with the locations where the respective signalsof subset A show activity at the respective time points.In other words, for the time points within the basic cyclelength where corresponding sensor data has been sam-pled, the size value for respective areas of active tissuecan be computed by adding up the size of surface ele-ments associated with the locations showing activity atthe respective time points. Alternatively, an average sur-face element size can be determined for a plurality ofsurface elements and the average size can be multipliedby the number of surface elements associated with therespective active area locations to assess the size of theareas of active tissue. It may also be possible that asmoothening operation, such as for example low-passfiltering, is performed on the resulting curve of area ofactive tissue (AAT) to smoothen it. Other smootheningalgorithms may be used instead. The selected area typ-ically corresponds to one or more atria of a patient butmay be modified by a medically trained person or by thesignal analyzer component. For a reentrant tachycardiamechanism, this area can e.g. be restricted to the pathof excitation, excluding the areas which merely representa dead end running into an anatomical obstacle.[0011] The system further has a visualizer componentconfigured to set a visual property for at least a subsetof the surface elements. The subset includes surface el-ements associated with locations identifying a particulararea of active tissue which is smaller than active tissueareas associated with other locations. In other words,tissue which is active when the AAT size value is minimalor at least smaller than the size value at other time points

3 4

EP 3 354 197 A1

4

5

10

15

20

25

30

35

40

45

50

55

can be highlighted. The visual property is selected to in-dicate, in a visualization of the selected area, the partic-ular area of active tissue as a potential ablation area tocure atrial flutter. It is a goal in atrial flutter treatment torestrict ablation to the smallest possible area which cancure the atrial flutter and optimizes cardiac function afterrestoring sinus rhythm by considering patient specific an-atomical variations. Thus, it is advantageous for the med-ically trained person to receive one or more suggestionsfrom the system for such potential ablation areas whichare smaller than other parts of the atrium affected by theatrial flutter activity. The medically trained person canthen make a decision for treatment of a particular areabased on this information. For example, in case of a focalsource atrial flutter the potential ablation area corre-sponds to a spot defining the approximate origin of thefocal source. In case of a micro-reentry atrial flutter thepotential ablation area corresponds to a bridging areabetween the approximate center of the micro-reentry af-fected area and the closest anatomical obstacle. In caseof a macro-reentry atrial flutter the potential ablation areacorresponds to the critical Isthmus within the macro-reen-try affected area.[0012] The visualization of the respective areas can beachieved by any suitable visual property setting for thecorresponding surface elements. For example, the visualproperty value may represent particular color for high-lighting the respective surface elements, a particular tex-ture for highlighting the respective surface elements, ora particular animation mode for highlighting the respec-tive surface elements.[0013] In one embodiment, the particular area of activetissue size is computed as the minimal area of activetissue at a respective time point within the basic cyclelength, the respective visual property value being indic-ative of the minimal area. In this embodiment, all surfaceelements associated with sensor locations of subset Ashowing activity at the point in time where the area ofactive tissue size reaches its minimum value during thebasic cycle length may receive the same indicative set-ting of the visual property (e.g., a particular color, a par-ticular pattern, a particular blinking mode, etc.). All othersurface elements receive visual property values differentfrom the indicative setting. In this embodiment only onepotential ablation area is indicated and therefore sug-gested to the medically trained person who needs finallyto decide whether the suggested indicated area is ap-propriate for the therapy.[0014] In some cases it may be advantageous topresent a plurality of alternative potential ablation areas.For example, there may be more than one area which issmall enough for a successful ablation operation. Ofcourse, only one of them is the smallest but the othersmay be equally suited for the intervention when consid-ering additional patient specific information. In this case,the medically trained person is better supported in herdecision making when multiple options of similar qualityare presented to select from. This is achieved by an em-

bodiment, where the particular area of active tissue iscomputed by computing statistical measures for the sizevalues of active tissue areas measured during the activityperiod of a respective signal. The respective visual prop-erty value is thereby indicative of the statistical measurevalue characterizing the active tissue area size value dur-ing the activity period. In other words, in this embodiment,for each signal location a visual property value is com-puted which represents the statistical measure of theAAT size values measured during the activity time inter-val. Statistical measures can be determined by variousstatistical methods, such as for example, arithmetic av-eraging, median, mode, or first 10-quantile. Other statis-tical methods may also be applicable. That is, small areasof active tissue which are not the smallest one are notfiltered out in this embodiment but visualized with a dif-ferent but similar visual property as the minimal area. Forexample, besides averaging, the sizes of active tissueareas measured during the activity period of a respectivesignal can also be assessed by computing the median,mode or quantile. In one embodiment, the 10% quantileof active tissue areas observed within the activity periodof a respective signal can be determined as visual prop-erty value. To avoid unnecessary visual coding (e.g.,color coding, pattern coding, etc.) for areas which are notof interest at all as a potential ablation area, only suchsurface elements may be visually coded which belong tosensor locations associated with areas of active tissuebelow a threshold value. Such a threshold value may bepredefined or may be computed at runtime, e.g. as apercentage of the maximal area of active tissue duringthe basic cycle length, or as a percentage surcharge onthe minimal area of active tissue.[0015] Current approaches to identify the critical partof the atrial flutter mechanism are based on visual in-spection. The LATs of recorded signals are compared tothe cumulative effect of atrial depolarization as indicatedby the P-wave from the body-surface electrocardiogram.Signals which are recorded at the critical site of the fluttercircuit are expected to show a LAT which is located tem-porarily between two consecutive P-waves. This meas-ure is chosen, as the P-wave represents the best avail-able surrogate for the atrial depolarization when no intra-atrial measurement data are available. In turn, the activetissue area as described herein is expected to be a morereliable indicator for the true amount of atrial depolariza-tion. From electromagnetic field theory, the P-wave isformed by projecting the electrical field vector as gener-ated by the atrial depolarization on the axes spanned bythe recorded electrodes of the surface ECG. Directly as-sessing the active atrial tissue avoids this projection. De-tails about this current approach can be found in theabove cited De Ponti reference.[0016] It may be advantageous for the decision supportof the medically trained person to know which atrial fluttertype is present in the atrium. Therefore, to solve this prob-lem, in one embodiment, the signal analyzer componentmay further determine, for each signal of at least a subset

5 6

EP 3 354 197 A1

5

5

10

15

20

25

30

35

40

45

50

55

B (which can be part of subset A) of the received signals,an active interval when electrical activity occurs, and aresting interval when no electrical activity occurs. Theanalyzer may then compute, based on the determinedintervals, a subset cycle length coverage as the ratio ofduration of activity (DoA) and the basic cycle lengthwherein the duration of activity includes the active inter-vals for the entire subset B of signals. Electrical activityhereby is a consequence of the depolarization of cardiactissue close to the sensor. No electrical activity occurswhen neighboring tissue is in rest and does not depolar-ize. Thus the electrical activity reflects changes of thetrans-membrane voltage of adjacent myocardial cells.For example, the selected area associated with subsetB may correspond to an atrium. The selection may bepredefined (e.g., the right atrium or the left atrium) or itmay be received by the system via an appropriate userinterface from a user of the system. For example, a phy-sician may use a pointing mechanism (e.g., by using amouse, a touch screen, or the like) to indicate selectedareas of interest in a graphical visualization of the pa-tient’s heart. In other words, the physician may selectboth atria, only one atrium, or just a part of an atrium asarea of interest for the later diagnostic analysis. Further,the signal analyzer may compute, based on the deter-mined intervals (active intervals and resting intervals), asubset cycle length coverage as the ratio of duration ofactivity and the basic cycle length. Thereby, the durationof activity includes the active intervals for the entire sub-set B of signals. In other words, the duration of activitycorresponds to one or more intervals which result fromsuperposing all active intervals detected by the furthersensors located within the selected area associated withsubset B. In this embodiment, the visualizer componentof the system can indicate to the medically trained humanthe selected area associated with subset B (e.g., right orleft atrium) as an area including a potential focal sourceif the subset cycle length coverage is smaller than a pre-defined ratio. If the subset cycle length coverage is equalor above the predefined ratio, the selected area is indi-cated as an area including a potential re-entry. The indi-cation may be implemented through visualization param-eters setting a predefined color or pattern for the visual-ization of the identified atrial flutter type. "Focal source"and "re-entry" are the major atrial flutter types as theyare known by the medically trained user of the system.For example, when a potential focal source is determinedin the selected area, the selected area can be visualizedto the medically trained person in a first color or using afirst pattern associated with the atrial flutter type "focalsource". When a potential re-entry is determined in theselected area, a second color or second pattern may beused which is associated with the atrial flutter type "re-entry".[0017] A first indicator for the nature of the underlyingtachycardia mechanism is given by the amount of cyclelength which can be annotated when all recorded signalsare assessed in conjunction. Reentry is typically suspect-

ed as mechanism for atrial flutter if activation times forat least 85% - 95% of the BCL (i.e., predefined ratio: 95%)can be successfully mapped. For each moment of thecovered time, accordingly, an EGM can be found some-where in the atrium which exhibits activity and whoselocal activity time can be annotated. Inversely, no activitycan be detected for less than 5% -15% of the cycle length.Given a sufficiently high-density of intra-cardiac meas-urements, the atrium is suspected to be in rest duringthis inactive time, excluding the presence of a re-entrantsource and endorsing a truly focal source. The coverageof macro re-entry activation was shown to be about95.964.3% (range 90% to 100 %) in clinical mappingstudies (cf., the above cited De Ponti reference). This isone important indicator that can be assessed when allrecorded electrograms are analyzed jointly.[0018] The buffer of 15% is frequently applied sincesome endocardial aspects of the flutter cycle may not bereachable during mapping, or the flutter may propagateon the epicardial aspect of the atrium. Thus the excitationwave cannot be observed at these locations and no LATcan be annotated. In addition, also highly fractionatedpotentials are assigned just one LAT value, despite thefact that they may cover significant parts of the cyclelength. Studies have shown prolonged activation at thecritical isthmus lasting mean durations of 200680 ms,with individual potentials having duration of 360 ms asillustrated in "Flutter localized to the anterior left atriumafter catheter ablation of atrial fibrillation (by Jais, P.;Sanders, P.; Hsu, L.-F.; Hocini, M.; Sacher, F.; Taka-hashi, Y.; Rotter, M.; Rostock, T.; Bordachar, P.; Reuter,S.; Laborderie, J.; Clementy, J.; Haissaguerre, M.; 2006;Journal of Cardiovascular Electrophysiology; 17;279-285)". This demonstrated a situation in which theannotation of LAT will not lead to optimal results. Instead,an activity based approach is suggested in the following.[0019] Although the annotation of LAT may not lead tooptimal results, in one embodiment, a predefined LATmay be used to define a surrogate for the activity. There-fore, a time window between 10 milliseconds and 300milliseconds (advantageously between 20ms and 30ms)around the LAT may be defined as active interval. Thismay be done independently of assessing the signal mor-phology, and thus not reflect the true activity. However,it may be used as an initial interval length which mayalready lead to results reflecting the general flutter typeor allowing to define active tissue areas. For example, inthis embodiment, the active interval (i.e. the predefinedwindow) may be centered around a particular point intime where the respective signal shows activity. The par-ticular time point can be the maximum or the minimumof the signal amplitude during a basic cycle length, or itmay be the maximum of the absolute value of the timederivative of the signal. Typically, the activity of the signalswings first to a maximum value of the signal amplitudeand then to a minimum value of the signal amplitude be-fore turning back to a value close to zero (resting interval).Between the time points with maximum and minimum

7 8

EP 3 354 197 A1

6

5

10

15

20

25

30

35

40

45

50

55

values of the signal amplitude the absolute value of thetime derivative reaches its maximum. The predefinedwindow may be centered around any one of these timepoints.[0020] In one embodiment, the system can provide ad-ditional information regarding the type of atrial flutter re-entry. This flutter type includes two sub-types: micro-re-entry and macro-re-entry. Both atrial flutter types requiredifferent treatment of the patient. Therefore, it is advan-tageous if the medically trained person is provided withthe additional information about the subtype of a potentialre-entry. In this embodiment, once a potential re-entry isdetected in the selected area, the signal analyzer com-ponent further can select a plurality of sub-areas withinthe originally selected area wherein the sub-areas havea size which is smaller than the size of the selected area,and wherein each sub-area has an overlapping area withat least one further sub-area. For example, the sub-areasmay have a circular shape or another geometric shape(e.g., square, rectangle, triangle, etc.) which is appropri-ate to sub-divide the selected area into evenly spreadsub-areas with overlapping areas between two neighbor-ing sub-areas.[0021] The signal analyzer can then compute for eachsub-area the respective sub-area cycle length coverage.This computation is performed in an analogous mannerto the computation of the sub-set cycle length coverage.However, instead of using the subset of signals originat-ing from locations within the selected area for determin-ing the active intervals and resting intervals only signalsoriginating from locations within the respective sub-areaare used for the computation. In other words, for the com-putation of a particular sub-area cycle length coveragethe corresponding ratio of duration of activity and the ba-sic cycle length takes into account the active intervals forthe signals originating from a location within the particularsub-area.[0022] Because of the higher spatial granularity of thesub-area cycle length coverages, there may be two ormore neighboring sub-areas with different sub-area cyclelength coverage values leading to a conflict situation ofthe respective overlapping area. To solve this problemwith this embodiment, for each overlapping area withconflicting sub-area cycle length coverages from at leasttwo overlapping sub-areas, one or more predefined con-flict resolution rules are applied to determine a respectiveoverlapping area cycle length coverage based on theconflicting sub-area cycle length coverages. For exam-ple, any one of the following conflict resolution rules maybe used by the signal analyzer: assigning the conflictingsub-area cycle length coverage with the highest value tothe respective overlapping area cycle length coverage;assigning the average value of the conflicting sub-areacycle length coverages to the respective overlapping ar-ea cycle length coverage; and assigning a weighted av-erage value of the conflicting sub-area cycle length cov-erages to the respective overlapping area cycle lengthcoverage.

[0023] In the case of the atrial flutter type "focal source"the duration of activity is approximately the same in allsub-areas. However, the origin of the focal source canbe determined to be in the sub-area where the activityoccurs first in time. In other words, the sub-area showingthe earliest activity can be proposed to the physician asa potential target for later ablation.[0024] Further, in this embodiment, the visualizer com-ponent can indicate sub-areas and overlapping areaswith respective cycle length coverages equal or abovean atrial flutter type threshold as areas including a po-tential micro re-entry, and further indicate sub-areas andoverlapping areas with respective cycle length coverag-es below the atrial flutter type threshold as areas not in-cluding a potential micro re-entry. If no sub-area or over-lapping area indicates the presence of a micro re-entry,a conclusion for the presence of a macro re-entry can bedrawn. The indication at the level of sub-areas may beimplemented through visualization parameters setting apredefined color or pattern associates with the respectiveatrial flutter type in a similar way as for the selected areas.[0025] The spatial resolution regarding the identifica-tion of potential locations comprising an atrial fluttersource depends on the size and shape of the sub-areas.In one embodiment, the sub-areas have a circular shapeand the overlapping areas between two overlapping cir-cles account for 1/5 to 1/3 of the sub-area diameter. Inan alternative embodiment, the sub-areas have a circularshape with a diameter between 2 cm and 4 cm, and thecircle centers of two neighboring sub-areas have a dis-tance between 0,5 cm and 1,5 cm. In another embodi-ment, the sub-areas have a square shape with a diagonalbetween 2 cm and 4 cm, and the square centers of twoneighboring sub-areas have a distance between 0,5 cmand 1,5 cm. In an alternative embodiment, the sub-areashave a square shape and the overlapping areas betweentwo overlapping circles account for 1/5 to 1/3 of the sub-area diagonal. A person skilled in the art can use othergeometric forms for the sub-areas and achieve a com-parable coverage of the selected area with a comparableresolution by applying similar overlapping ratios as dis-closed for circle and square shaped sub-areas.[0026] In one embodiment, computing a subset cyclelength coverage may include applying a Boolean OR op-erator to the determined activity and resting intervals ofthe respective signals so that any activity interval of anysignal contributes to the duration of activity, and no ac-tivity in all signals contributes to the duration of resting.[0027] In one embodiment, a computer-implementedmethod is provided for supporting diagnostic analysis ofpotential ablation areas to cure atrial flutter. The methodmay be executed by the previously disclosed system.The method includes: receiving an electrical referencesignal generated by a reference sensor, the referencesignal providing a time reference for electrical activity ofone or more atria of a patient wherein the time intervalbetween two subsequent electrical activities is referredto as basic cycle length; receiving a plurality of electrical

9 10

EP 3 354 197 A1

7

5

10

15

20

25

30

35

40

45

50

55

signals generated by one or more further sensors where-in the plurality of electrical signals relate to a plurality oflocations in the one or more atria, the plurality of locationscomprising locations different from the location of the ref-erence sensor; determining, for each signal of at least asubset of the received signals originating from locationswithin a selected area of the one or more atria, whenelectrical activity occurs; determining, for a plurality oftime points within the basic cycle length, size values forrespective areas of active tissue by assessing the aggre-gate size of surface elements associated with the loca-tions where the respective signals show activity at therespective time points; setting a visual property for atleast a subset of the surface elements wherein the subsetincludes surface elements associated with locationsidentifying a particular area of active tissue which is small-er than active tissue areas associated with other loca-tions, to indicate, in a visualization of the selected area,the particular area of active tissue as a potential ablationarea to cure atrial flutter.[0028] In one embodiment, in case of a focal sourceatrial flutter, the potential ablation area may be visualizedby highlighting a spot defining the approximate origin ofthe focal source. In one embodiment, in case of a micro-reentry atrial flutter the potential ablation area may behighlighted as a bridging area between the approximatecenter of the micro-reentry affected area and the closestanatomical obstacle. In one embodiment, in case of amacro-reentry atrial flutter the potential ablation area maybe highlighted as the critical isthmus within the macro-reentry affected area.[0029] In one embodiment, the method may further in-clude computing the size value of the particular area ofactive tissue as the minimal area of active tissue at arespective time point within the basic cycle length where-in the respective visual property value is selected so thatit is indicative of the minimal area (e.g., a particular colordistinguishing the minimal area from the rest of the se-lected area.[0030] In one embodiment, the method further includescomputing the size value of the particular area of activetissue as statistical measure for the sizes of active tissueareas measured during the activity period of a respectivesignal, the respective visual property value being indic-ative of the statistical measure value. The statisticalmeasure may be determined by statistical methods, suchas for example, arithmetic averaging, median, mode, orfirst 10-quantile.[0031] In one embodiment, the method may further in-clude steps to provide suggestions for an atrial fluttertype by: determining, for each of the signals of at leastthe subset, an active interval when electrical activity oc-curs, and a resting interval when no electrical activityoccurs; based on the determined intervals, computing asubset cycle length coverage as the ratio of duration ofactivity (DoA) and the basic cycle length (BCL) whereinthe duration of activity (DoA) includes the active intervalsfor the entire subset of signals; if the subset cycle length

coverage is smaller than a predefined ratio then indicat-ing the selected area as an area including a potentialfocal source, else indicating the selected area as an areaincluding a potential re-entry.[0032] In one embodiment, a computer program prod-uct is provided for supporting diagnostic analysis of po-tential ablation areas to cure atrial flutter. The computerprogram product is loaded into a memory of a computingdevice/system as disclosed herein and executed by atleast one processor of the computing device. This causesthe computing device to execute the steps of the abovedisclosed computer-implemented method thereby imple-menting the features of the system as disclosed herein.[0033] In one embodiment, the method may further in-clude: if the selected area includes a potential re-entry,selecting a plurality of sub-areas within the originally se-lected area wherein the sub-areas have a size which issmaller than the size of the selected area, and whereineach sub-area has an overlapping area with at least onefurther sub-area; computing for each sub-area the re-spective sub-area cycle length coverage; for each over-lapping area with conflicting sub-area cycle length cov-erages from at least two overlapping sub-areas, applyingone or more predefined conflict resolution rules to deter-mine a respective overlapping area cycle length cover-age based on the conflicting sub-area cycle length cov-erages; indicating sub-areas and overlapping areas withrespective cycle length coverages equal to or above anatrial flutter type threshold as areas including a potentialmicro re-entry. Optionally, the method may include indi-cating sub-areas and overlapping areas with respectivecycle length coverages below the atrial flutter type thresh-old as areas including a potential macro re-entry. Thepresence of a macro re-entry can be concluded if no sub-area or overlapping area indicates the presence of a mi-cro re-entry. As previously disclosed, the indication ofthe various atrial flutter types can be controlled by cor-responding visualization parameter associated with therespective atrial flutter types. The visualization parame-ters may represent predefined different colors, patterns,animations, image depth in 3D visualizations, etc. for thevarious flutter types. When the system visualizes a par-ticular atrial flutter type, the associated visualization pa-rameter is used to render the respective area (selectedarea or sub-areas) in a graphical representation of thepatient’s atria.[0034] For example, a predefined conflict resolutionrule can include assigning the conflicting sub-area cyclelength coverage with the highest value to the respectiveoverlapping area cycle length coverage. In this example,the visualization parameter for the respective overlap-ping area equals the visualization parameter of the cor-responding sub-area with the highest value.[0035] For example, a predefined conflict resolutionrule can include assigning the average value of the con-flicting sub-area cycle length coverages to the respectiveoverlapping area cycle length coverage. In this example,a visualization parameter for the overlapping area is com-

11 12

EP 3 354 197 A1

8

5

10

15

20

25

30

35

40

45

50

55

puted based on visualization parameters of the sub-ar-eas having conflicting cycle length coverages. For exam-ple, if the visualization parameter of a first sub-area in-dicates a red visualization and the visualization param-eter of a second sub-area indicates a blue visualization,the visualization parameter of the overlapping area mayresult in a violet visualization. The average value may becomputed based on RAL values of the respective colors.Other number values describing colors may be used in-stead. In case of using pattern based visualization pa-rameters, in one example the first sub-area may havethe visualization parameter horizontal hatch and the sec-ond sub-area may have a vertical hatch pattern. The av-erage may be computed as the overlay of both hatchpatterns resulting in a cross hatch. In the case of usinganimations (e.g., blinking of the respective area) the an-imation frequency may be the visualization parameter foraveraging.[0036] For example, a predefined conflict resolutionrule can include assigning a weighted average value ofthe conflicting sub-area cycle length coverages to therespective overlapping area cycle length coverage. Aperson skilled in the art can adjust the conflict resolutionrules for assigning average values by simply addingweighting factors which may emphasize, for example,the visualization parameter of the sub-area having thesub-area cycle length coverage with the highest value.[0037] Further aspects of the invention will be realizedand attained by means of the elements and combinationsparticularly depicted in the appended claims. It is to beunderstood that both, the foregoing general descriptionand the following detailed description are exemplary andexplanatory only and are not restrictive of the inventionas described.

Brief Description of the Drawings

[0038]

FIG. 1 is a simplified block diagram illustrating anembodiment of a decision support computer systemfor supporting diagnostic analysis of potential abla-tion areas to cure atrial flutter;FIG. 2 is a simplified flowchart of a computer-imple-mented method for supporting diagnostic analysisof potential ablation areas to cure atrial flutter ac-cording to an embodiment of the invention;FIG. 3 is a schematic illustration of a human heartwith the right atrium showing a micro-reentry atrialflutter;FIG. 4 is a schematic illustration of an atrium withrepresentations of sensor locations;FIG. 5 illustrates sensor signals received from differ-ent locations of a patient’s atrium with activity inter-vals;FIG. 6A is a graph illustrating the area of active tissueover time during the basic cycle length with averagedactive tissue area values for the active periods of two

sensors;FIG. 6B is a graph illustrating the area of active tissueover time with the minimal area of active tissue duringthe basic cycle length;FIG. 7 is a schematic illustration of an atrium withrepresentations of sensor locations and an indicationof a potential ablation area;FIG. 8 shows a magnified portion of FIG. 7 includingthe potential ablation area with highlighted surfaceelements;FIG. 9 shows two examples of active tissue graphsfor a right and a left atrium based on real sensor data;FIG. 10 shows an example of an atrium which is colorcoded according to one embodiment to indicate po-tential ablation areas;FIG. 11 illustrates sensor signals received from dif-ferent locations of a patient’s atrium with aggregateactivity periods during the basic cycle length;FIG. 12 illustrates the duration of activity and theduration of resting for the determination of cyclelength coverage;FIG. 13 is a simplified illustration of overlapping sub-areas on a graphic representation of an atrium ac-cording to one embodiment of the invention; andFIG. 14 is a diagram that shows an example of ageneric computer device and a generic mobile com-puter device which may be used with the techniquesdescribed herein.

Detailed Description

[0039] FIG. 1 is a simplified block diagram illustratingan embodiment of a decision support computer system100 for supporting diagnostic analysis of potential abla-tion areas to cure atrial flutter. FIG. 2 is a simplified flow-chart of a computer-implemented method 1000 for sup-porting diagnostic analysis of potential ablation areas tocure atrial flutter according to an embodiment of the in-vention. The functions of system 100 of FIG. 1 are dis-cussed in the context of the method steps of method 1000which are performed by the respective system compo-nents of system 100. Therefore, the following descriptionrefers to reference numbers of the FIGs. 1 and 2.[0040] The system 100 includes an interface compo-nent 110 configured to receive 1100, 1200 data 240 fromone or more external data sources 200. The external datasources may be sensors CSCS, S1 to Sn providing realtime data about the electric activation of a patient’s atriaor it may be a data storage device 210 which provideshistoric (previously recorded or simulated) sensor dataabout the electric activation of the patient’s atria.[0041] The received data 240 includes an electrical ref-erence signal RS generated by a reference sensor. Typ-ically, a coronary sinus catheter sensor CSCS is used tomeasure the electric activity in a patient’s atria at thelocation of the coronary sinus. This reference signal RSprovides a position reference and a time reference forelectrical activity of one or more atria of the patient.

13 14

EP 3 354 197 A1

9

5

10

15

20

25

30

35

40

45

50

55

Thereby, the time interval between two subsequent elec-trical activities is referred to as basic cycle length (BCL).[0042] The interface further receives 1200 a pluralityof electrical signals P1 to Pi generated by one or morefurther sensors S1 to Sn wherein the plurality of electricalsignals P1 to Pi relate to a plurality of locations in the oneor more atria, the plurality of locations being different fromthe location of the reference sensor. Typically, such sen-sors are multi-polar mapping catheters which are widelyused for electro-anatomical mapping systems EAMS.Examples of such sensors are described in detail in thereferences: "A multi-purpose spiral high-density mappingcatheter: initial clinical experience in complex atrial ar-rhythmias; (by Jones, D. G.; McCready, J. W.; Kaba, R.A.; Ahsan, S. Y.; Lyne, J. C.; Wang, J.; Segal, O. R.;Markides, V.; Lambiase, P. D.; Wong, T.; Chow, A. W.C.; 2011; Journal of Interventional Cardiac Electrophys-iology: an International Journal of Arrhythmias and Pac-ing; 31: 225-235)", and "Rapid high resolution electroan-atomical mapping: evaluation of a new system in a canineatrial linear lesion model; (by Nakagawa, H.; Ikeda, A.;Sharma, T.; Lazzara, R.; Jackman, W. M.; 2012; Circu-lation. Arrhythmia and Electrophysiology; 5; 417-424)".[0043] The system 100 further has a signal analyzercomponent 120. The signal analyzer 120 determines, foreach signal P1, P2, P3 of at least a subset 121 of thereceived signals P1 to Pi when an electrical activity oc-curs at the location of the respective sensor. Thereby,the subset 121 includes such signals P1, P2, P3 whichoriginate from locations within a selected area of the oneor more atria. Typically, the selected area is one atrium(of the patient’s heart) which is subject to the diagnosticanalysis. The selected area may also relate to a portionof an atrium. The selection of the selected area may bereceived from the medically trained person performingthe diagnosis or it may be predefined in the system 100.For example, in a first analysis the right atrium (or a por-tion) may be selected and in a second analysis the leftatrium (or a portion) may be selected. Based on the ac-tivity intervals of the various signals in the subset 121,for each signal in the subset 121, the signal analyzerdetermines 1300 an active interval when electrical activ-ity occurs. For example, the detection of activity may beperformed by assessing the amplitude of the measuredsignal in that a window is shifted over the signal with themaximum and minimum amplitude being computed andthe current window being identified as activity during anactive interval if the difference between maximum andminimum values exceeds a certain threshold. In an al-ternative implementation, the detection of activity maybe performed by assessing the instantaneous energy ofthe measured signal, in which a threshold can be usedto distinguish active and inactive parts of the signal. Thedurations of the electrical activity of the signals P1, P2,P3 in subset 121 are illustrated by the rectangular shapeswhich are superposed to the signal curves.[0044] An active area component 122 of the signal an-alyzer 120 determines 1400, for a plurality of time points

within the basic cycle length BCL, respective areas ofactive tissue AAT by aggregating the size of surface el-ements (e.g., by adding the size of the surface elements,counting the number of surface elements and multiply byan average size, etc.) associated with the locationswhere the respective signals show activity at the respec-tive time points. FIG. 8 discusses in detail how surfaceelements can be associated with sensor locations. TheAAT graph illustrated in FIG. 1 shows an example em-bodiment where a minimal area of active tissue MAAT isdetermined. This embodiment is discussed in detail inthe context of FIG. 6B.[0045] By using the above disclosed signal analysisthe system 100 can compute one or more areas in thepatient’s atrium which are appropriate for a potential ab-lation operation. Such one or more potential ablation ar-eas are then conveyed by the system 100 to the medicallytrained person to support the diagnosis and take a deci-sion for the appropriate treatment of the patient. For thispurpose, the system includes a visualizer component 130which sets 1500 a visual property (e.g., a visual propertyselected from the earlier disclosed examples, such as aparticular color for highlighting the respective surface el-ements, a particular texture for highlighting the respectivesurface elements, or a particular animation mode forhighlighting the respective surface elements) for at leasta subset of the surface elements. The subset includessurface elements associated with locations identifying aparticular area of active tissue which is smaller than ac-tive tissue areas associated with other locations whichare active at other time points. The particular area of ac-tive tissue is then indicated, in a visualization of the se-lected area, as a potential ablation area to cure atrialflutter. As explained earlier, it can be advantageous tohighlight not only the smallest (minimal) area of activetissue but also other areas of active tissue which may bebigger but similar in size to the smallest area, and there-fore, may represent potential ablation areas of similarquality.[0046] As discussed above, visual properties, such ascolor, patterns or animations may be used to render agraphic representation 230 of the patient’s atria on a dis-play device 220 (e.g., communicatively coupled with thesystem 100 via interface 110) highlighting one or moreareas of active tissue as potential ablation areas in sucha way that the medically trained user of the system isprovided with clear information about alternative optionsfor ablation treatment. Any appropriate display technol-ogy (e.g., monitor, touch screen, etc.) may be used forthe visualization of the potential ablation area(s) 231.[0047] For example, in case of a focal source atrialflutter the potential ablation area typically correspondsto a spot defining the approximate origin of the focalsource. For example, in case of a micro-reentry atrialflutter the potential ablation area corresponds to a bridg-ing area between the approximate center of the micro-reentry affected area and the closest anatomical obsta-cle. For example, in case of a macro-reentry atrial flutter

15 16

EP 3 354 197 A1

10

5

10

15

20

25

30

35

40

45

50

55

the potential ablation area corresponds to the critical isth-mus within the macro-reentry affected area.[0048] FIG. 3 is a schematic illustration 300 of a humanheart with the right atrium RA showing a micro-reentryatrial flutter 310. The human heart includes the right andleft atria RA, LA and the right and left ventricles RV, LV.The illustration also shows the aorta AO and the pulmo-nary aorta PA. In the right atrium RA the three arrows310 representing a circular excitation wave illustrate thepresence of a micro-reentry in the patient’s right atriumRA.[0049] FIG. 4 is a schematic illustration of an atrium500 with representations of sensor locations S1, S2, S3,S4, etc. For convenience of illustration only the sensorlocations S1 to S4 have a reference sign in the figure.The atrium 500 can be a right or a left atrium. The sensorlocations are illustrated by black dots. The sensor loca-tions do not necessarily correspond to locations wheresensors are operated simultaneously. One sensor mayhave been moved to various sensor locations for takingmeasurements regarding the electrical activity at the re-spective location. In real world measurement scenariosup to about 12000 sensor locations may provide a highspatial resolution image of the electrical activity in theatrium. Measurements at different locations can be takenat different times but may be merged afterwards whencomputing the areas of active tissue because the elec-trical activation occurs periodically with a period lengthcorresponding to the basic cycle length. Therefore, onlythe relative times of the activity intervals of the sensorsignals within a basic cycle length are relevant for thedetermination of the potential ablation areas.[0050] FIG. 5 illustrates sensor signals RS, P1 to P4received from different locations of a patient’s atrium withactivity intervals reflecting activation for particular loca-tions of the atrium. In the example, the reference signalRS is shown at the top providing the time reference forthe signals P1 to P4 which are taken by sensors S1 toS4 (cf. FIG. 4) at locations different from the location ofthe reference sensor (not shown in FIG. 4). The referencesignal RS illustrates the periodic activity observed at thecoronary sinus. The time between two voltage amplitudepeaks is defined as the basic cycle length BCL. Besidesmaxima in the voltage, also other characteristic morpho-logical markers of the electrical signal, such as for exam-ple, minima, strongest gradient, and so forth can be usedto assign the reference time. As the reference signal cor-responds to an electrical signal which is generated by ahuman organ and not by a clock signal of a technicalsystem, there may be small variability in the BCL lengthover time. This has been discussed in the Saoudi refer-ence cited above. Continuously monitoring the CS cath-eter reference signal allows to compensate for such smallvariability in BCL by scaling the duration of the resultingactive and inactive periods. For example, every 300 mil-liseconds a variance of +/- 15 milliseconds may occur(modulation of BCL caused by respiration). By comparingthe activity intervals with the corresponding actual BCL,

the computed subset cycle length coverage is automat-ically normalized with regards to the actual BCL. Theactual BCL can be manually defined by the medicallytrained person or can be based on a median value of allmeasured actual BCL or it may be based on a meanvalue of all measured actual BCL.[0051] The signal analyzer determines for each signalP1 to P4 the activity intervals for each cycle. In the ex-ample of FIG. 5 a cycle is indicated as the time betweentwo of the vertical dotted lines. However, it is not relevantwhere a cycle starts and ends as long as the cycle lengthcorresponds to BCL. The activity intervals of the respec-tive signals are illustrated by the rectangular overlays.For simplicity, only the rectangular overlay 301 is shownwith a reference number. The length of the resting intervalof a cycle is defined as the difference between BCL andthe length of the respective active interval of the signal.[0052] In the example of FIG. 4 it is assumed that theatrium 500 (cf. FIG. 4) suffers from a macro-reentry flutterwhich causes an excitation wave 520 from the right tothe left. Such an excitation wave of a macro-reentry istypically rotating around the opening 510 of the atrium(to the corresponding ventricle). That is, the excitationwave moves from left to right at the backside of the atrium(illustrated by dashed arrow 520) to re-appear at the rightside on the front side (the illustrated view). Turning backto FIG. 5, sensor S4 measures the activity in P4 shortlybefore sensor S1 measures activity in P1. During a firstoverlapping time interval including time point T1 S1 andS4 measure activity simultaneously. When the wavepropagates to the left, sensor S3 measures the activityin P3 shortly before sensor S2 measures the activity inP2. During a second overlapping time interval includingtime point T2 S3 and S2 measure activity simultaneously.[0053] FIG. 6A shows a graph 610 illustrating the sizeof the area of active tissue AAT over time t during thebasic cycle length BCL with averaged active tissue areasize values AAT(P1), AAT(P4) for the active periods ofthe two sensors S1, S4, respectively. The overlappingactivity intervals of P4 and P1 are again illustrated assubsequent rectangular shapes which include the timepoint T1. The AAT 610 shows a minimum at T1 duringthe BCL. The time points T0, T2, T3 are associated withlarger areas of active tissue than T1. For example, T2corresponds to the time point included in the overlap ofthe activity intervals (i.e. activity periods) of P2, P3. Inthis embodiment, the size of the particular area of activetissue is computed as statistical measure (in the exam-ple, the average values AAT(P4), AAT(P1) of the AATvalues during the activity intervals of P4, P1) for the sizesof active tissue areas measured during the activity periodof the respective signals. The respective visual propertyvalues are indicative of the statistical measure valuesAAT(P4), AAT(P1).[0054] In other words, the AAT size values around thetime point T1 are significantly smaller than for other timepoints during the BCL. This means that the size of thepotential ablation area at locations which are active

17 18

EP 3 354 197 A1

11

5

10

15

20

25

30

35

40

45

50

55

around T1 is relatively small compared to the other areasaffected by the macro-reentry. To visualize this informa-tion for diagnosis support to the medically trained userthe visualizer component is setting visual property valuesfor the surface elements being associated with the sen-sors S1 and S4 which reflect the averaged AAT valuesfor the corresponding activity intervals. As a result, thesurface elements associated with S4 receive a visualproperty value reflecting AAT(P4) and the surface ele-ments associated with S1 receive a visual property valuereflecting AAT(P1). For example, a color scheme maybe used, where the color spectrum from red to blue isdistributed over the AAT values of the curve 610. Thecolor value at T3 (AAT curve maximum) may be assignedto blue and the color value at T1 (AAT curve minimum)may be assigned to red. The color assigned to the aver-aged AAT value AAT(P4) would then be in the red partof the spectrum with a slight shift towards orange. Thecolor for AAT(P1) would also be in the red part of thespectrum where the shift into orange exceeds the one ofAAT(P4). The red colored area is then an indicator for apotential ablation area. The curve 610 may show multiplelocal minima dependent on the patient’s heart disease.In such cases, there may be multiple areas with visualproperty values in the red part of the spectrum indicatingvarious alternatives for the ablation therapy. The medi-cally trained person finally decides where to apply theablation therapy.[0055] Instead of averaging, other statistical methods(e.g., median, mode, first 10-quantile, etc.) may be usedto determine the statistical measures for the activity pe-riods at the respective sensor locations. Further, othervisual properties for highlighting may be used as de-scribed earlier.[0056] In one embodiment, a threshold value may beset for the statistical measure values. In this embodiment,only the surface elements of sensor locations are high-lighted as potential ablation areas where the statisticalmeasure value of the AAT sizes is below the thresholdvalue. All other surface elements may be associated witha default visual property value (e.g., color=blue). Thisembodiment provides an improved distinction of potentialablation areas from other areas.[0057] FIG. 6B is a graph 620 illustrating the size ofarea of active tissue AAT over time with the minimal areaof active tissue MAAT during the basic cycle length BCL.In this embodiment, only the MAAT is indicated to themedically trained person. The advantage over the em-bodiment of FIG. 6A is that the smallest possible areafor potential ablation is indicated. The disadvantage isthat only one area is indicated and no alternatives arehighlighted. In this embodiment, the minimum value ofthe AAT curve 620 can be determined by any algorithmfor the detection of minima. In the example, the minimalvalue MAAT is found at time point T1. In case of usingagain color as the visual property, the color values of allsurface elements associated with sensor locations show-ing activity at T1 (in the example: S1, S4) are set to the

color value (e.g., red) reflecting the smallest AAT valueMAAT during the BCL. The visual property values of allother surface elements may be set to a different color(e.g., blue). For example, the default color value of allsurface elements may be blue. The default value is thenoverwritten with the MAAT color value once identified bythe analyzer and visualizer components. A person skilledin the art can use other visual properties to achieve asimilar highlighting effect for the medically trained user.[0058] FIG. 7 is a schematic illustration of the atrium500 with representations of sensor locations as black bul-let points and representations of the sensor locations ofS1, S4 as circles. Around S1, S4 a potential ablation area501 is indicated. In the example, the excitation wavepropagating through the front side of the atrium is illus-trated through the area between the dashed lines 502,503. The dashed lines may correspond to the boundaryto previously ablated or scarred tissue, which limits thearea of excitable atrial tissue and prevents the atrial ex-citation from leaving the dashed boundary. The minimalarea of active tissue is at the highlighted sensor locationsS1, S4 which corresponds, in the example, to the criticalisthmus. The shape of potential ablation area 501 is aschematic illustration. In a real implementation, theshape is determined by the surface texture used for a 3Dvisualization of the atrium 500. This is explained in detailin FIG. 8.[0059] FIG. 8 shows a magnified portion of FIG. 7 in-cluding the potential ablation area with highlighted sur-face elements. Typically, 3D visualizations of three-di-mensional objects use surface textures based on poly-gons. In the example, the surface of the 3D visualizationof the atrium 500 is based on triangles 1 - 29. Other pol-ygons can be used as well by persons skilled in the art.In the example of FIG. 8 a pattern property is used asvisual property. The surface elements associated withthe sensor locations S1, S4 show a dotted pattern fillingwhereas the other surface elements have a white filling.Once the sensor locations of S1, S4 are determined asthe locations associated with the minimal area of activetissue (MAAT embodiment), the visual property valuesof the surface elements associated with the sensor loca-tions are set to the dotted pattern filling.[0060] A person skilled in the art may use various al-gorithms to uniquely assign each surface element to ex-actly one sensor location. In the example, the shortestdistance of the center of gravity of a triangle surface el-ement to the neighboring sensor locations determines towhich sensor location the surface element belongs. Inthe example, the surface elements 1, 2, 15, 17, 18, 19,25 are assigned to S1. The surface elements 5, 6, 8, 9,10, 13 are assigned to S4. Instead of evaluating the short-est distance to the center of gravity, other methods (e.g.,distance to closest vertices of the polygon, number ofclosest polygon vertices, etc.) may be used to assign thesurface elements to sensor locations.[0061] In the MAAT embodiment, both sensors showsimultaneous activity during the time point T1 (cf. FIG.

19 20

EP 3 354 197 A1

12

5

10

15

20

25

30

35

40

45

50

55

6B) when the MMAT value is reached and the surfaceelements of both sensor locations receive the same vis-ual property value (dotted pattern filling) reflecting theMMAT value at the time point T1. In the embodimentusing statistical measures, the surface elements as-signed to S1 receive a visual first property value (e.g., afirst color) which is different from a second visual propertyvalue of the surface elements assigned to S4 (e.g., asecond color). Thereby, the first visual property valuereflects the statistical AAT measure computed for the ac-tivity interval of S1 and the second visual property valuereflects the statistical AAT measure computed for the ac-tivity interval of S4.[0062] FIG. 9 shows two examples of active tissuegraphs for a right and a left atrium RA, LA based on realsensor data. The sensor data was sampled at a samplingrate of 2035 Hz. The time axis in the figure has the unitof samples received. The basic cycle length in the exam-ple is approximately at 660 samples (corresponding toapproximately 330 ms). The area of active tissue dimen-sion shows the fraction of active tissue compared to thetotal area of tissue (i.e., the sum of all surface elementsrepresenting the surface of the atrium). For the right atri-um RA this fraction reaches almost 60 % whereas for theleft atrium LA the fraction only reaches almost 40 %. Be-tween 100 samples (corresponding to 50 ms) and 230samples (corresponding to 115 ms), the AAT curve forthe RA has the value of 0, indicating that no activity wasfound in this atrium for the corresponding approximately20% of BCL. Thus the presence of a reentrant mecha-nism can be excluded for the RA. In the LA curve for AAT,the curve does fall below 0.03, indicating that activity waspresent in the LA at each point in time of the BCL. Thisfavors the presence of a reentrant atrial flutter mecha-nism which is located in the LA and passively causes theRA activation. The fraction visualization of the AAT curveas shown in FIG. 9 is equivalent to the absolute area sizevisualization as used in FIGs. 6A, 6B. Both implementa-tions may be used with every embodiment.[0063] Looking at the right graph for the left atrium LA,the graph shows three local minima but does not showa AAT value of zero at any time point during the BCL. Inthe embodiment identifying the MMAT only the minimalarea of active tissue is indicated. That is, the surfaceelements associated with sensor locations showing ac-tivity at the point in time when the MMAT is active (ap-proximately at 590 samples) receive a visual propertyvalue which indicates the MMAT. The other local minimaat approximately 110 samples and 260 samples are nothighlighted in this embodiment. Nevertheless, at leastthe area of active tissue associated with the first localminimum (110 samples) may be equally suitable as theMMAT.[0064] When using the embodiment based on statisti-cal measures at least the first local minimum (110 sam-ples) is also highlighted with a visual property value sim-ilar to the one computed for the absolute minimum. Alsothe AAT associated with the second local minimum (260

samples) may be highlighted with a visual property valuewhich indicates that the respective AAT size is largerthan the AAT size associated with the other minima but,nevertheless, significantly smaller than the AAT size as-sociated with time points between 300 samples and 500samples. In this embodiment, a threshold value may beused to cut off the visualization of AATs above the thresh-old value. For example, a threshold value of 0.075 wouldcut off the second minimum (260 samples) and only theAATs associated with the first and third minima would behighlighted as potential ablation areas.[0065] FIG. 10 shows an example of an atrium visual-ization 600 which is color coded with a grey scale 610according to one embodiment to indicate potential abla-tion areas. The visualization is based on the embodimentusing statistical measures where the surface elementsassociated with respective sensors receive visual prop-erty values indicating the statistical measure value (e.g.,average value) for the AAT during the active period ofthe respective sensor signal. The grey scale values re-flect fraction AAT values from 0 to 0.1. When inspectingthe complete curve of fraction AAT values (not shown),all values were found to be above 0 for each time stepwithin the BCL, indicating a reentrant atrial flutter mech-anism. Manual visual inspection of the LAT map of thispatient (not shown) confirmed the reentrant mechanismand revealed a zone of slow conduction in this atrium,located at the anterior wall and being in agreement withthe area of FIG. 10 colored in bright grey scale values.This confirmed the role of the area showing low fractionAAT values as critical isthmus for this patient. In the ex-ample, in the upper part, two areas 621, 622 with lowfraction AAT values are highlighted indicating potentialablation areas which have a comparable size smallerthan the other color coded areas. Dependent on the typeof the atrial flutter the medically trained person can thenmake a decision about the location for the ablation ther-apy. The atrial flutter type can be derived from inspectionof the LAT map by a medically trained person, or an in-dication of the atrial flutter type may be derived by furtherfunctions of the computer system which are explained inthe following.[0066] FIG. 11 illustrates sensor signals P1’ to P4’ re-ceived from different locations of a patient’s atrium withaggregate activity periods ai1 to ai8 during the basic cyclelength BCL. The sensor signals P1’ to P4’ may be iden-tical to the signals P1 to P4 of FIG. 5. In this embodiment,the computer system of FIG. 1 can implement furtherfunctions to assist the medically trained person to recog-nize which type of atrial flutter is present in the patient’satrium.[0067] In this embodiment, the signal analyzer compo-nent 120 (cf. FIG. 1) may determine, for each signal ofat least the subset of the received signals originating fromlocations within the selected area, an active interval ai1to ai8 when electrical activity occurs, and a resting inter-val when no electrical activity occurs. Based on the de-termined intervals, a subset cycle length coverage can

21 22

EP 3 354 197 A1

13

5

10

15

20

25

30

35

40

45

50

55

be computed as the ratio of duration of activity (DoA) andthe basic cycle length (BCL) wherein the duration of ac-tivity includes the active intervals for the entire subset ofsignals. The visualizer component 130 (cf. FIG. 1) canindicate the selected area as an area including a potentialfocal source if the subset cycle length coverage is smallerthan a predefined ratio. Otherwise, the selected area isindicated as an area including a potential re-entry.[0068] For computing the subset cycle length cover-age, the analyzer component firstly determines the du-ration of activity for the active intervals for the entire sub-set of signals P1’ to P4’. For example, a Boolean ORoperator can be applied to the rectangular shapes (e.g.,333) representing the lengths of the determined intervalsof respective signals so that any activity interval of thesignals of the subset contributes to the duration of activity.To determine the overall duration of resting (across thesignals of the subset). With regards to the duration ofresting, only periods with no activity in all of the signalscan contribute when a Boolean OR operator is applied.The bottom line of FIG. 11 illustrates the result of the ORoperator applied to the signals P1’ to P4’. In each cycletwo activity intervals are detected which are illustratedby the dashed frames. In the first cycle, the Boolean ORapplication results in the activity intervals ai1, ai2. In thesecond cycle the result is ai3, ai4, and so forth.[0069] Based on the determined intervals, a compara-tor component of the signal analyzer 120 computes asubset cycle length coverage as the ratio of duration ofactivity (DoA) and the basic cycle length BCL whereinthe duration of activity includes the active intervals forthe entire subset 121 of signals. In other words, the DoAcorresponds to the total active time of the atrium in theselected area which is monitored by the sensors provid-ing the signals of the subset 121. The comparator com-pares the determined subset cycle length coverage to apredefined ratio. According to studies (e.g., the study il-lustrated in the above cited reference "A multi-purposespiral high-density mapping catheter..." by Jones et al.)it is known that the predefined ratio which can be usedto distinguish the selected area with regards to the atrialflutter types focal source and re-entry is approximately85%-95% (the exact number may vary according to dif-ferent studies). If the selected area is an area which in-cludes a potential focal source then the subset cyclelength coverage determined from the signals of the sub-set 121 is smaller than said predefined ratio. If the subsetcycle length coverage is equal or above said predefinedratio, the selected area is an area including a potentialre-entry.[0070] As discussed above, visual parameters, suchas color, patterns or animations may be used to rendera graphic representation of the patient’s active atria tis-sue on a display device in such a way that the medicallytrained user of the system is provided with clear informa-tion about the potential clinical state (focal source vs. re-entry) of the patient’s atria which facilitates the diagnosticanalysis by the user. Any appropriate display technology

(e.g., monitor, touch screen, etc.) may be used for thevisualization of the atrial flutter types.[0071] In one embodiment, an additional filter functionmay be used by the signal analyzer to discard signalsfrom the subset which may be tampered with pulses fromthe patient’s heart’s ventricles. In this embodiment, a sur-face electrogram ECG can also be recorded simultane-ously to the intracardiac electrical signals P1 to Pi, andcan then be analyzed to detect ventricular depolariza-tions (QRS complexes). Intracardiac signals may becompromised by far fields originating from ventricular de-polarizations. This may lead to electrical activity in theintracardiac signal causing deviations from baseline andthus potentially being misinterpreted as activity originat-ing from an atrial source. Thus the activity of the atriumshould be assessed while no QRS is present. The filterfunction may now discard or ignore any signals of thesubset during such intervals where QRS is present.[0072] FIG. 12 illustrates the duration of activity DoAand the duration of resting DoR for the determination ofcycle length coverage. The determined activity intervalsai1, ai2 of the first cycle are added to provide the overallduration of activity DoA. By computing the ratio betweenthe duration of DoA and BCL the subset cycle lengthcoverage is computed for the selected area which is mon-itored by the subset of signals P1’ to P4’. It is to be notedthat in a real world diagnosis scenario the number of sig-nals in a subset can easily exceed 1000 in cases wherea database is providing historic sensor data. In real timemonitoring scenarios the number of simultaneously re-ceived signals is limited by the number of sensors whichcan be operated simultaneously in the patient’s heart. Itis to be noted that historic sensor data, as used herein,may have been collected very recently (e.g., a few sec-onds or minutes before being processed by the signalanalyzer). That is, although not representing real-timedata, historic sensor data may also reflect the currentmedical state of the patient’s atria.[0073] FIG. 13 is a simplified example of overlappingsub-areas sa1 to sa4 on a graphic representation 500(corresponding to the selected area) of an atrium accord-ing to an embodiment of the invention.[0074] Upon detection of a potential re-entry in the se-lected area (i.e., If the subset cycle length coverage isdetermined equal or above the predefined ratio), the sig-nal analyzer component further selects a plurality of sub-areas sa1 to sa4 within the originally selected area 500.The sub-areas are automatically selected by the signalanalyzer component according to predefined selectionrules so that each sub-area has a size which is smallerthan the size of the selected area 500, and each sub-area has an overlapping area oa12 to oa34 with at leastone further sub-area. The predefined selection rules mayinclude a predefined shape for sub-areas and predefinedoverlapping parameters. For example, in one embodi-ment, the selection rules may define the sub-areas ashaving a circular shape with a diameter between 2 and4 cm. The sub-area dimensions relate to the respective

23 24

EP 3 354 197 A1

14

5

10

15

20

25

30

35

40

45

50

55

heart portion covered by a particular sub-area (and notto the dimensions of the visualization of the heart whichmay be magnified or demagnified). The overlapping pa-rameters may define that the overlapping areas betweentwo overlapping circles account for 1/5 to 1/3 of the sub-area diameter. In an alternative embodiment, the over-lapping parameters may define circle centers of twoneighboring sub-areas having a distance between 0,5cm and 1,5 cm. In another embodiment, the shape pa-rameters may define a square shape with a diagonal be-tween 2 cm and 4 cm, and the overlapping parametersmay define square centers of two neighboring sub-areashaving a distance between 0,5 cm and 1,5 cm. In analternative embodiment, the sub-areas have a squareshape and the overlapping areas between two overlap-ping circles account for 1/5 to 1/3 of the sub-area diag-onal. A person skilled in the art can use other geometricforms for the sub-areas and achieve a comparable cov-erage of the selected area with a comparable resolutionby applying similar overlapping ratios as disclosed forcircle and square shaped sub-areas. Further, the selec-tion rules may define particular portions of the selectedarea as areas of interest with regards the analysis of atrialflutter types. Such areas of interest may be received fromthe medically trained user or they may be predefined inthe system. The signal analyzer can then distribute thesub-areas according the corresponding selection rulesacross the entire selected area or across one or moreparticular portions (areas of interest) of the selected area.[0075] In the example of FIG. 13, for the reason of sim-plicity, only one row of overlapping sub-areas sa1 to sa2with circle shapes is shown are shown as overlays to therepresentation 700 of an atrium. The skilled person, how-ever, will understand that further rows may exist aboveand below the shown row which have further overlappingareas with the sub-areas sa1 to sa4. The selection rulesin this example can define such further rows of overlap-ping sub-areas in such a way that the circle centers wouldbe positioned on a vertical line defined by the intersectionpoints of two neighboring sub-areas. In this example,each overlapping area has overlap portions where threesub-areas are overlapping. In another implementation,the selection rules can define such further rows of over-lapping sub-areas in such a way that the circle centerswould be positioned on a vertical line intersecting withthe circle center of the respective above or below sub-area. In this implementation the overlapping area mayrelate to two or four sub-areas depending on the sizeparameters of the respective sub-areas. A person skilledin the art can define other appropriate selection rules toachieve sufficient coverage of the entire selected areaor selected areas of interest with appropriate overlappingsub-areas.[0076] The signal analyzer can then compute the re-spective sub-area cycle length coverage for each sub-area sa1 to sa4. The computation may be performed bythe same algorithm which is used for the previous com-putation of the subset cycle length coverage for the se-

lected areas. However, only the signals originating fromsensors at locations covered by the respective sub-areaare used for the computation of the respective sub-areacycle length coverage.[0077] It may occur that different sub-area cycle lengthcoverage values are computed for neighboring sub-are-as. In this case, there is a conflict with regards to thecycle length coverage to be assigned to the overlappingareas oa12, oa23, oa34 which is defined as the intersec-tion of the overlapping sub-areas sa1-sa2, sa2-sa3, andsa3-sa4, respectively. To resolve this conflict, for eachoverlapping area with conflicting sub-area cycle lengthcoverages from at least two overlapping sub-areas, thesignal analyzer applies one or more predefined conflictresolution rules to determine a respective overlappingarea cycle length coverage based on the conflicting sub-area cycle length coverages. Different approaches forconflict resolution can be defined by different conflict res-olution rules. For example, the signal analyzer may as-sign the conflicting sub-area cycle length coverage withthe highest value to the respective overlapping area cyclelength coverage as a high sensitivity to high cycle lengthcoverage may be desired to identify potential micro-reen-try atrial flutter mechanisms. In another example, it mayassign the average value of the conflicting sub-area cyclelength coverage values to the respective overlapping ar-ea cycle length coverage. In a further example, it mayassign a weighted average value of the conflicting sub-area cycle length coverage values to the respective over-lapping area cycle length coverage.[0078] Once the sub-area cycle length coverage val-ues and overlapping area cycle length coverage valuesare computed, the signal analyzer compares the com-puted values to an atrial flutter type threshold which maybe different from the predefined ratio used by the firstcomparison for distinguishing between the atrial fluttertypes focal source and re-entry. However, the atrial fluttertype threshold used by the second comparison may alsobe equal or similar to the predefined ratio. The atrial fluttertype threshold is chosen as a threshold value which canbe used to distinguish between the atrial flutter sub-types"micro-re-entry" and "macro-re-entry". Different medicalstudies propose different threshold values for this pur-pose. For example a threshold of 75% of BCL coveredby the DoA within a sub-area of 2cm diameter is proposedin the previously cited work "A deductive mapping strat-egy for atrial tachycardia following atrial fibrillation abla-tion: importance of localized reentry" by Jais et al.. Acycle length coverage value below the atrial flutter typethreshold is seen as an indicator for a potential "macro-re-entry". A cycle length coverage value equal to or abovethe atrial flutter type threshold is seen as an indicator fora potential "micro-re-entry".[0079] The signal analyzer provides the result of thecomparison regarding each sub-area sa1 to sa4 andeach overlapping area oa12, oa23, oa34 to the visualizercomponent which can render a visualization of the sub-areas and overlapping areas accordingly. The rendering

25 26

EP 3 354 197 A1

15

5

10

15

20

25

30

35

40

45

50

55

is performed in such a way that it indicates to the medi-cally trained person a potential micro-re-entry if the re-spective cycle length coverage value is equal to or abovethe atrial flutter type threshold. Further, the renderingmay indicate a potential macro-re-entry if the respectivecycle length coverage value is below the atrial flutter typethreshold (as described above). The indication is imple-mented via similar visualization parameters as disclosedalready above for the rendering of the selected area. Thatis, different colors, patterns, animations, 3D-effects, etc.may be used to graphically distinguish areas with poten-tial micro-re-entries from areas with potential macro-re-entries for the medically trained user to support the di-agnostic decision making.[0080] FIG. 14 is a diagram that shows an example ofa generic computer device 900 and a generic mobilecomputer device 950, which may be used with the tech-niques described here. Computing device 900 is intend-ed to represent various forms of digital computers, suchas laptops, desktops, workstations, personal digital as-sistants, tablets, servers, blade servers, mainframes,and other appropriate computers. Generic computer de-vice 900 may correspond to a computer system 100 asillustrated in FIG. 1. Computing device 950 is intendedto represent various forms of mobile devices, such aspersonal digital assistants, cellular telephones, smartphones, and other similar computing devices. For exam-ple, computing device 950 may be used by a user as afront end to interact with the computer system 100. Com-puting device may, for example, include the display de-vice 220 of FIG. 1. The components shown here, theirconnections and relationships, and their functions, aremeant to be exemplary only, and are not meant to limitimplementations of the inventions described and/orclaimed in this document.[0081] Computing device 900 includes a processor902, memory 904, a storage device 906, a high-speedinterface 908 connecting to memory 904 and high-speedexpansion ports 910, and a low speed interface 912 con-necting to low speed bus 914 and storage device 906.Each of the components 902, 904, 906, 908, 910, and912, are interconnected using various busses, and maybe mounted on a common motherboard or in other man-ners as appropriate. The processor 902 can process in-structions for execution within the computing device 900,including instructions stored in the memory 904 or on thestorage device 906 to display graphical information for aGUI on an external input/output device, such as display916 coupled to high speed interface 908. In other imple-mentations, multiple processing units and/or multiplebuses may be used, as appropriate, along with multiplememories and types of memory. Also, multiple computingdevices 900 may be connected, with each device provid-ing portions of the necessary operations (e.g., as a serverbank, a group of blade servers, or a processing device).[0082] The memory 904 stores information within thecomputing device 900. In one implementation, the mem-ory 904 is a volatile memory unit or units. In another im-

plementation, the memory 904 is a non-volatile memoryunit or units. The memory 904 may also be another formof computer-readable medium, such as a magnetic oroptical disk.[0083] The storage device 906 is capable of providingmass storage for the computing device 900. In one im-plementation, the storage device 906 may be or containa computer-readable medium, such as a floppy disk de-vice, a hard disk device, an optical disk device, or a tapedevice, a flash memory or other similar solid state mem-ory device, or an array of devices, including devices in astorage area network or other configurations. A computerprogram product can be tangibly embodied in an infor-mation carrier. The computer program product may alsocontain instructions that, when executed, perform one ormore methods, such as those described above. The in-formation carrier is a computer- or machine-readable me-dium, such as the memory 904, the storage device 906,or memory on processor 902.[0084] The high speed controller 908 manages band-width-intensive operations for the computing device 900,while the low speed controller 912 manages lower band-width-intensive operations. Such allocation of functionsis exemplary only. In one implementation, the high-speedcontroller 908 is coupled to memory 904, display 916(e.g., through a graphics processor or accelerator), andto high-speed expansion ports 910, which may acceptvarious expansion cards (not shown). In the implemen-tation, low-speed controller 912 is coupled to storage de-vice 906 and low-speed expansion port 914. The low-speed expansion port, which may include various com-munication ports (e.g., USB, Bluetooth, Ethernet, wire-less Ethernet) may be coupled to one or more input/out-put devices, such as a keyboard, a pointing device, ascanner, or a networking device such as a switch or rout-er, e.g., through a network adapter.[0085] The computing device 900 may be implement-ed in a number of different forms, as shown in the figure.For example, it may be implemented as a standard server920, or multiple times in a group of such servers. It mayalso be implemented as part of a rack server system 924.In addition, it may be implemented in a personal computersuch as a laptop computer 922. Alternatively, compo-nents from computing device 900 may be combined withother components in a mobile device (not shown), suchas device 950. Each of such devices may contain one ormore of computing device 900, 950, and an entire systemmay be made up of multiple computing devices 900, 950communicating with each other.[0086] Computing device 950 includes a processor952, memory 964, an input/output device such as a dis-play 954, a communication interface 966, and a trans-ceiver 968, among other components. The device 950may also be provided with a storage device, such as amicrodrive or other device, to provide additional storage.Each of the components 950, 952, 964, 954, 966, and968, are interconnected using various buses, and severalof the components may be mounted on a common moth-

27 28

EP 3 354 197 A1

16

5

10

15

20

25

30

35

40

45

50

55

erboard or in other manners as appropriate.[0087] The processor 952 can execute instructionswithin the computing device 950, including instructionsstored in the memory 964. The processor may be imple-mented as a chipset of chips that include separate andmultiple analog and digital processing units. The proces-sor may provide, for example, for coordination of the othercomponents of the device 950, such as control of userinterfaces, applications run by device 950, and wirelesscommunication by device 950.[0088] Processor 952 may communicate with a userthrough control interface 958 and display interface 956coupled to a display 954. The display 954 may be, forexample, a TFT LCD (Thin-Film-Transistor Liquid CrystalDisplay) or an OLED (Organic Light Emitting Diode) dis-play, or other appropriate display technology. The displayinterface 956 may comprise appropriate circuitry for driv-ing the display 954 to present graphical and other infor-mation to a user. The control interface 958 may receivecommands from a user and convert them for submissionto the processor 952. In addition, an external interface962 may be provided in communication with processor952, so as to enable near area communication of device950 with other devices. External interface 962 may pro-vide, for example, for wired communication in some im-plementations, or for wireless communication in otherimplementations, and multiple interfaces may also beused.[0089] The memory 964 stores information within thecomputing device 950. The memory 964 can be imple-mented as one or more of a computer-readable mediumor media, a volatile memory unit or units, or a non-volatilememory unit or units. Expansion memory 984 may alsobe provided and connected to device 950 through ex-pansion interface 982, which may include, for example,a SIMM (Single In Line Memory Module) card interface.Such expansion memory 984 may provide extra storagespace for device 950, or may also store applications orother information for device 950. Specifically, expansionmemory 984 may include instructions to carry out or sup-plement the processes described above, and may in-clude secure information also. Thus, for example, expan-sion memory 984 may act as a security module for device950, and may be programmed with instructions that per-mit secure use of device 950. In addition, secure appli-cations may be provided via the SIMM cards, along withadditional information, such as placing the identifying in-formation on the SIMM card in a non-hackable manner.[0090] The memory may include, for example, flashmemory and/or NVRAM memory, as discussed below.In one implementation, a computer program product istangibly embodied in an information carrier. The compu-ter program product contains instructions that, when ex-ecuted, perform one or more methods, such as thosedescribed above. The information carrier is a computer-or machine-readable medium, such as the memory 964,expansion memory 984, or memory on processor 952,that may be received, for example, over transceiver 968

or external interface 962.[0091] Device 950 may communicate wirelesslythrough communication interface 966, which may includedigital signal processing circuitry where necessary. Com-munication interface 966 may provide for communica-tions under various modes or protocols, such as GSMvoice calls, SMS, EMS, or MMS messaging, CDMA, TD-MA, PDC, WCDMA, CDMA2000, or GPRS, EDGE,UMTS, LTE, among others. Such communication mayoccur, for example, through radio-frequency transceiver968. In addition, short-range communication may occur,such as using a Bluetooth, WiFi, or other such transceiver(not shown). In addition, GPS (Global Positioning Sys-tem) receiver module 980 may provide additional navi-gation- and location-related wireless data to device 950,which may be used as appropriate by applications run-ning on device 950.[0092] Device 950 may also communicate audibly us-ing audio codec 960, which may receive spoken infor-mation from a user and convert it to usable digital infor-mation. Audio codec 960 may likewise generate audiblesound for a user, such as through a speaker, e.g., in ahandset of device 950. Such sound may include soundfrom voice telephone calls, may include recorded sound(e.g., voice messages, music files, etc.) and may alsoinclude sound generated by applications operating ondevice 950.[0093] The computing device 950 may be implement-ed in a number of different forms, as shown in the figure.For example, it may be implemented as a cellular tele-phone 980. It may also be implemented as part of a smartphone 982, personal digital assistant, or other similar mo-bile device.[0094] Various implementations of the systems andtechniques described here can be realized in digital elec-tronic circuitry, integrated circuitry, specially designedASICs (application specific integrated circuits), computerhardware, firmware, software, and/or combinationsthereof. These various implementations can include im-plementation in one or more computer programs that areexecutable and/or interpretable on a programmable sys-tem including at least one programmable processor,which may be special or general purpose, coupled toreceive data and instructions from, and to transmit dataand instructions to, a storage system, at least one inputdevice, and at least one output device.[0095] These computer programs (also known as pro-grams, software, software applications or code) includemachine instructions for a programmable processor, andcan be implemented in a high-level procedural and/orobject-oriented programming language, and/or in as-sembly/machine language. As used herein, the terms"machine-readable medium" and "computer-readablemedium" refer to any computer program product, appa-ratus and/or device (e.g., magnetic discs, optical disks,memory, Programmable Logic Devices (PLDs)) used toprovide machine instructions and/or data to a program-mable processor, including a machine-readable medium

29 30

EP 3 354 197 A1

17

5

10

15

20

25

30

35

40

45

50

55

that receives machine instructions as a machine-reada-ble signal. The term "machine-readable signal" refers toany signal used to provide machine instructions and/ordata to a programmable processor.[0096] To provide for interaction with a user, the sys-tems and techniques described here can be implementedon a computer having a display device (e.g., a CRT (cath-ode ray tube) or LCD (liquid crystal display) monitor) fordisplaying information to the user and a keyboard and apointing device (e.g., a mouse or a trackball) by whichthe user can provide input to the computer. Other kindsof devices can be used to provide for interaction with auser as well; for example, feedback provided to the usercan be any form of sensory feedback (e.g., visual feed-back, auditory feedback, or tactile feedback); and inputfrom the user can be received in any form, includingacoustic, speech, or tactile input.[0097] The systems and techniques described herecan be implemented in a computing device that includesa backend component (e.g., as a data server), or thatincludes a middleware component (e.g., an applicationserver), or that includes a front end component (e.g., aclient computer having a graphical user interface or aWeb browser through which a user can interact with animplementation of the systems and techniques describedhere), or any combination of such backend, middleware,or frontend components. The components of the systemcan be interconnected by any form or medium of digitaldata communication (e.g., a communication network).Examples of communication networks include a local ar-ea network ("LAN"), a wireless local area network("WLAN"), a wide area network ("WAN"), and the Inter-net.[0098] The computing device can include clients andservers. A client and server are generally remote fromeach other and typically interact through a communica-tion network. The relationship of client and server arisesby virtue of computer programs running on the respectivecomputers and having a client-server relationship to eachother.[0099] A number of embodiments have been de-scribed. Nevertheless, it will be understood that variousmodifications may be made without departing from thespirit and scope of the invention.[0100] In addition, the logic flows depicted in the figuresdo not require the particular order shown, or sequentialorder, to achieve desirable results. In addition, othersteps may be provided, or steps may be eliminated, fromthe described flows, and other components may be add-ed to, or removed from, the described systems. Accord-ingly, other embodiments are within the scope of the fol-lowing claims.

Claims

1. A decision support computer system (100) for sup-porting diagnostic analysis of potential ablation are-

as to cure atrial flutter, comprising:

an interface component (110) configured to re-ceive an electrical reference signal (RS) gener-ated by a reference sensor (CSCS), the refer-ence signal (RS) providing a time reference forelectrical activity of one or more atria of a patientwherein the time interval between two subse-quent electrical activities is referred to as basiccycle length (BCL), and further configured to re-ceive a plurality of electrical signals (P1 to Pi)generated by one or more further sensors (S1to Sn) wherein the plurality of electrical signals(P1 to Pi) relate to a plurality of locations in theone or more atria, the plurality of locations com-prising locations different from the location ofthe reference sensor (CSCS);a signal analyzer component (120) configuredto determine, for each signal of at least a subsetof the received signals originating from locationswithin a selected area of the one or more atria,when electrical activity occurs, and to deter-mine, for a plurality of time points within the basiccycle length, size values for respective areas ofactive tissue by assessing the aggregate size ofsurface elements associated with the locationswhere the respective signals show activity at therespective time points; anda visualizer component (130) configured to seta visual property value for at least a subset ofthe surface elements wherein the subset in-cludes surface elements associated with loca-tions identifying a particular area of active tissuewhich is smaller than active tissue areas asso-ciated with other locations, the visual propertyvalue indicating, in a visualization of the selectedarea, the particular area of active tissue as apotential ablation area to cure atrial flutter.

2. The system of claim 1, wherein in case of a focalsource atrial flutter the potential ablation area corre-sponds to a spot defining the approximate origin ofthe focal source, or in case of a micro-reentry atrialflutter the potential ablation area corresponds to abridging area between the approximate center of themicro-reentry affected area and the closest anatom-ical obstacle, or in case of a macro-reentry atrial flut-ter the potential ablation area corresponds to the crit-ical isthmus within the macro-reentry affected area.

3. The system of claim 1 or 2, wherein the selectedarea corresponds to an atrium.

4. The system of any of previous claims, wherein thevisual property is selected from the group of: a par-ticular color for highlighting the respective surfaceelements, a particular texture for highlighting the re-spective surface elements, and a particular anima-

31 32

EP 3 354 197 A1

18

5

10

15

20

25

30

35

40

45

50

55

tion mode for highlighting the respective surface el-ements.

5. The system of any of the previous claims, whereinthe particular area of active tissue is computed asthe minimal area of active tissue at a respective timepoint within the basic cycle length, the respectivevisual property value being indicative of the minimalarea.

6. The system of any of the claims 1 to 4, wherein theparticular area of active tissue is computed by as astatistical measure for the sizes of active tissue areasmeasured during the activity period of a respectivesignal, the respective visual property value being in-dicative of the statistical measure value.

7. The system of any of the previous claims, wherein

the signal analyzer component (120) is furtherconfigured to determine, for each signal of atleast the subset of the received signals originat-ing from locations within the selected area (500),an active interval (ai1 to ai8) when electrical ac-tivity occurs, and a resting interval when no elec-trical activity occurs; and configured to compute,based on the determined intervals, a subset cy-cle length coverage as the ratio of duration ofactivity (DoA) and the basic cycle length (BCL)wherein the duration of activity includes the ac-tive intervals for the entire subset of signals; andthe visualizer component (130) further config-ured to indicate the selected area as an areaincluding a potential focal source if the subsetcycle length coverage is smaller than a prede-fined ratio, and else to indicate the selected areaas an area including a potential re-entry.

8. A computer-implemented method (1000) for sup-porting diagnostic analysis of potential ablation are-as to cure atrial flutter, the method comprising:

receiving (1100) an electrical reference signalgenerated by a reference sensor, the referencesignal providing a time reference for electricalactivity of one or more atria of a patient whereinthe time interval between two subsequent elec-trical activities is referred to as basic cyclelength;receiving (1200) a plurality of electrical signalsgenerated by one or more further sensorswherein the plurality of electrical signals relateto a plurality of locations in the one or more atria,the plurality of locations comprising locations dif-ferent from the location of the reference sensor;determining (1300), for each signal of at least asubset of the received signals originating fromlocations within a selected area of the one or

more atria, when electrical activity occurs;determining (1400), for a plurality of time pointswithin the basic cycle length, size values for re-spective areas of active tissue by assessing theaggregate size of surface elements associatedwith the locations where the respective signalsshow activity at the respective time points;setting (1500) a visual property value for at leasta subset of the surface elements wherein thesubset includes surface elements associatedwith locations defining a particular area of activetissue which is smaller than active tissue areasassociated with other locations, to indicate, in avisualization of the selected area, the particulararea of active tissue as a potential ablation areato cure atrial flutter.

9. The method of claim 8, wherein in case of a focalsource atrial flutter the potential ablation area corre-sponds to a spot defining the approximate origin ofthe focal source, or in case of a micro-reentry atrialflutter the potential ablation area corresponds to abridging area between the approximate center of themicro-reentry affected area and the closest anatom-ical obstacle, or in case of a macro-reentry atrial flut-ter the potential ablation area corresponds to the crit-ical isthmus within the macro-reentry affected area.

10. The method of any of the claims 8 to 9, wherein thevisual property is selected from the group of: a par-ticular color for highlighting the respective surfaceelements, a particular texture for highlighting the re-spective surface elements, and a particular anima-tion mode for highlighting the respective surface el-ements.

11. The method of any of the claims 8 to 10, wherein theparticular area of active tissue is computed as theminimal area of active tissue at a respective timepoint within the basic cycle length, the respectivevisual property value being indicative of the minimalarea.

12. The method of any of the claims 8 to 10, wherein theparticular area of active tissue is computed as sta-tistical measure for the sizes of active tissue areasmeasured during the activity period of a respectivesignal, the respective visual property value being in-dicative of the statistical measure value.

13. The method of claim 12, wherein the statistical meas-ure is determined by any one statistical method ofthe group of: arithmetic averaging, median, mode,and first 10-quantile.

14. The method of any of the claims 8 to 13, further com-prising:

33 34

EP 3 354 197 A1

19

5

10

15

20

25

30

35

40

45

50

55

determining, for each of the signals of at leastthe subset, an active interval when electrical ac-tivity occurs, and a resting interval when no elec-trical activity occurs;based on the determined intervals, computing asubset cycle length coverage as the ratio of du-ration of activity (DoA) and the basic cycle length(BCL) wherein the duration of activity (DoA) in-cludes the active intervals for the entire subsetof signals;if the subset cycle length coverage is smallerthan a predefined ratio then indicating the se-lected area as an area including a potential focalsource, else indicating the selected area as anarea including a potential re-entry.

15. A computer program product for supporting diagnos-tic analysis of potential ablation areas to cure atrialflutter, the computer program product when loadedinto a memory of a computing device and executedby at least one processor of the computing deviceexecutes the steps of the computer-implementedmethod according to any one of the claims 8 to 14.

35 36

EP 3 354 197 A1

20

EP 3 354 197 A1

21

EP 3 354 197 A1

22

EP 3 354 197 A1

23

EP 3 354 197 A1

24

EP 3 354 197 A1

25

EP 3 354 197 A1

26

EP 3 354 197 A1

27

EP 3 354 197 A1

28

EP 3 354 197 A1

29

EP 3 354 197 A1

30

EP 3 354 197 A1

31

5

10

15

20

25

30

35

40

45

50

55

EP 3 354 197 A1

32

5

10

15

20

25

30

35

40

45

50

55

EP 3 354 197 A1

33

5

10

15

20

25

30

35

40

45

50

55

EP 3 354 197 A1

34

REFERENCES CITED IN THE DESCRIPTION

This list of references cited by the applicant is for the reader’s convenience only. It does not form part of the Europeanpatent document. Even though great care has been taken in compiling the references, errors or omissions cannot beexcluded and the EPO disclaims all liability in this regard.

Non-patent literature cited in the description

• SAOUDI, N. ; COSIO, F. ; WALDO, A. ; CHEN, S.A. ; LESAKA, Y. ; LESH, M. ; SAKSENA, S. ;SALERNO, J. ; SCHOELS, W. A classification of atri-al flutter and regular atrial tachycardia according toelectrophysiological mechanisms and anatomicalbases; a Statement from a Joint Expert Group fromThe Working Group of Arrhythmias of the EuropeanSociety of Cardiology and the North American Soci-ety of Pacing and Electrophysiology. European HeartJournal, 2001, vol. 22, 1162-1182 [0002]

• PATEL, A. M. ; D’AVILA, A. ; NEUZIL, P. ; KIM, S.J. ; MELA, T. ; SINGH, J. P. ; RUSKIN, J. N. ; RED-DY, V. Y. Circulation. Arrhythmia and Electrophysi-ology, 2008, vol. 1, 14-22 [0002]

• SCHAEFFER, B. ; HOFFMANN, B. A. ; MEYER,C. ; ; AKBULAK, R. O. ; MOSER, J. ; JULARIC, M. ;EICKHOLT, C. ; NUHRICH, J. M. ; KUKLIK, P. ;WILLEMS, S. Characterization, Mapping and Abla-tion of Complex Atrial Tachycardia: Initial Experiencewith a Novel Method of Ultra High-Density 3D Map-ping. Journal of Cardiovascular Electrophysiology,2016, vol. 27, 1139-1150 [0002]

• JAIS, P. ; MATSUO, S. ; KNECHT, S. ; WEERA-SOORIYA, R. ; HOCINI, M. ; SACHER, F. ;WRIGHT, M. ; NAULT, I ; LELLOUCHE, N. ; KLEIN,G. A deductive mapping strategy for atrial tachycardiafollowing atrial fibrillation ablation: importance of lo-calized reentry. Journal of Cardiovascular Electro-physiology, 2009, vol. 20, 480-491 [0004]

• DE PONTI, R. ; VERLATO, R. ; BERTAGLIA, E. ;DEL GRECO, M. ; FUSCO, A. ; BOTTONI, N. ; DRA-GO, F.; ; SCIARRA, L. ; OMETTO, R. ; MAN-TOVAN, R. Treatment of macro-re-entrant atrial tach-ycardia based on electroanatomic mapping: identifi-cation and ablation of the mid-diastolic isthmus.Europace : European Pacing, Arrhythmias, and Car-diac Electrophysiology : Journal of the WorkingGroups on Cardiac Pacing, Arrhythmias, and CardiacCellular Electrophysiology of the European Societyof Cardiology, 2007, vol. 9, 449-457 [0004]

• D. G. LATCU ; S. - S. BUN ; F. VIERA ; T. DELASSI ;M. EL JAMILI ; A. AL AMOURA ; N. SAOUDI. Se-lection of Critical Isthmus in Scar-Related Atrial Tach-ycardia Using a New Automated Ultrahigh ResolutionMapping System. Circulation. Arrhythmia and Elec-trophysiology, 2017, vol. 10 [0004]

• JAIS, P. ; SANDERS, P. ; HSU, L.-F. ; HOCINI, M. ;SACHER, F. ; TAKAHASHI, Y. ; ROTTER, M. ;ROSTOCK, T. ; BORDACHAR, P. ; REUTER, S.Flutter localized to the anterior left atrium after cath-eter ablation of atrial fibrillation. Journal of Cardiovas-cular Electrophysiology, 2006, vol. 17, 279-285[0018]

• JONES, D. G. ; MCCREADY, J. W. ; KABA, R. A. ;AHSAN, S. Y. ; LYNE, J. C. ; WANG, J. ; SEGAL,O. R. ; MARKIDES, V. ; LAMBIASE, P. D. ; WONG,T. A multi-purpose spiral high-density mapping cath-eter: initial clinical experience in complex atrial ar-rhythmias. Journal of Interventional Cardiac Electro-physiology: an International Journal of Arrhythmiasand Pacing, 2011, vol. 31, 225-235 [0042]

• NAKAGAWA, H. ; IKEDA, A. ; SHARMA, T. ; LAZ-ZARA, R. ; JACKMAN, W. M. Rapid high resolutionelectroanatomical mapping: evaluation of a new sys-tem in a canine atrial linear lesion model. Circulation.Arrhythmia and Electrophysiology, 2012, vol. 5,417-424 [0042]