Research Article Ubiquitous Health Management System with …downloads.hindawi.com/journals/jam/2014/878741.pdf · 2019-07-31 · paper, we develop an ubiquitous health management

Research ArticleUbiquitous Health Management System with Watch-TypeMonitoring Device for Dementia Patients

Dongmin Shin Dongil Shin and Dongkyoo Shin

Department of Computer Engineering Sejong University 98 Gunja-Dong Gwangjin-Gu Seoul 143-747 Republic of Korea

Correspondence should be addressed to Dongkyoo Shin shindksejongackr

Received 11 November 2013 Revised 13 January 2014 Accepted 19 January 2014 Published 4 March 2014

Academic Editor Young-Sik Jeong

Copyright copy 2014 Dongmin Shin et alThis is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

For patients who have a senile mental disorder such as dementia the quantity of exercise and amount of sunlight are an importantclue for doses and treatmentTherefore monitoring daily health information is necessary for patientsrsquo safety and health A portableand wearable sensor device and server configuration for monitoring data are needed to provide these services for patients Awatch-type device (smart watch) that patients wear and a server system are developed in this paper The smart watch developedincludes a GPS accelerometer and illumination sensor and can obtain real time health information by measuring the positionof patients quantity of exercise and amount of sunlight The server system includes the sensor data analysis algorithm and webserver used by the doctor and protector to monitor the sensor data acquired from the smart watch The proposed data analysisalgorithm acquires the exercise information and detects the step count in patientsrsquo motion acquired from the acceleration sensorand verifies the three cases of fast pace slow pace and walking pace showing 96 of the experimental results If developed and theu-Healthcare System for dementia patients is applied higher quality medical services can be provided to patients

1 Introduction

The increase in the elderly population due to the developmentof medical technology is creating challenges for care profes-sionals and developers of ubiquitous healthcare systems

Dementia refers to the cognitive impairment usuallyaffecting old people and makes functioning in daily life moredifficult Early symptoms of dementia include memory lossgradually affecting everyday activities Typically from a fewmonths to several years the first symptoms are mild butdevelop slowly and gradually lead to serious memory lossIn addition dementia patients have difficulty in recognizingtheir family members and doing complicated tasks Theyusually have wandering symptoms and more than 73experience being lost or missing [1]

The ubiquitous healthcare system is a convergence ofinformation communication technology and healthcare andhas emerged in various ways to help these kinds of patients[2] Keruve a Spanish company provides a medical servicefor dementia patients This service uses a bracelet with abuilt-in GPS and a portable deviceThe GPS bracelet featuresprecise location detection using triangulation even if the

patient is in the room [3] Korea Telecom a Korean com-pany has developed a location-tracking system using GPSand Code Division Multiple Access (CDMA) [4] GangnamDistrict Office in Seoul Korea has developed a system calledGangnamU-Safe System [5]This service began in May 2009using Ubiquitous Sensor Network (USN) technology andGPS This system provides a compact device featured withan emergency alarm service used for the safety of sociallyvulnerable individuals including children and those withintellectual disabilities

Currently healthcare systems for patients with dementiaare focusing on location tracking using a Global PositioningSystem (GPS) For patients with mental disorders momen-tum monitoring and medical service profiling can managetheir risks and enhance their quality of life [6 7] In thispaper we develop an ubiquitous health management systemfor dementia patients to improve their health and safetyfollowing the concept Internet of Things (IoT) [8ndash10] Thesystem consists of a wrist watch-type device and a serversystem The device includes a built-in GPS ambient lightsensor and acceleration sensor and communicates withthe server system The server system functions include the

Hindawi Publishing CorporationJournal of Applied MathematicsVolume 2014 Article ID 878741 8 pageshttpdxdoiorg1011552014878741

2 Journal of Applied Mathematics

creation of a personal profile for patients and monitoringa patientrsquos location and measuring the amount of sunlightillumination and walking step count to use as medical dataThe system helps dementia patients avoid the risk of beingmissing or lost by wandering symptoms

2 Related Works

Recently the concept of Internet of Things (IoT) has beenapplied in ubiquitous healthcare systems and services [8ndash10] IoT is a novel paradigm of technologies that intercon-nect everyday objects with each other through the Internetexploiting multiple wireless communication interfaces andadvancements in computing devices [11] With the spread ofsmart phones and tablets loaded with various sensors such asGPS and accelerometers higher quality services are providedto the users by connection of the information on theWeb andreal world [12]

With the advent of IoT research on numerous medicalservices for patients has been performed [9 10] Researchon wireless networking technologies for developing a mobilehealthcare environment has been carried out and it leads intothe concept of mobile IoT (m-IoT) which is a new healthcareconnectivity paradigm that interconnects IP-based commu-nication technologies such as IPv6 over low power WPAN(6LoWPAN) with emerging 4G networks for future Internet-based healthcare services [9] Typically healthcare servicesare comprised of the sensors acquiring biosignals and theservers processing the huge amount of biodata generatedfrom the sensors Service platforms that interconnect cloudcomputing distributed processing and high speed dataprocessing systems following the concept of IoT are beingresearched for efficient healthcare services [10]

Studies on human movement detection and behavioralpatterns have been carried out in various ways for health-care services The motion recognition algorithm based ona motion-tree is developed using the acceleration featuresof a mobile phone [12] The motion detection algorithmis one of the basic methods for detecting the number ofwalking steps [13 14] Human movements are distinguishedby a pattern recognition algorithm and a way of extractingvarious motions are developed from basic motion patternsand feature vectors of humans This function reads normaland abnormalmovements for example sitting standing andfalling down as well as the number of steps [15ndash18]

Position tracking using GPS is one of the data for mea-suring the momentum as well as the current position of thepatient in a healthcare system Recently research on indoorposition tracking methods using Wi-Fi or other positioningschemes are being carried out because it is impossible to geta GPS signal indoors [19 20]

3 Development of a Ubiquitous HealthManagement System

The system consists of a watch-type monitoring deviceand server The monitoring device includes a GPS 3-axisaccelerometer and ambient light sensor It is worn on

the patientrsquos wrists and periodically transfers his activityinformation to the server derived from his location andamount of light illumination detecting sun exposure Thencare professionals and doctors can monitor the patientrsquoshealth condition through thewebpage delivered by the serverThe server identifies the location through the patientrsquos datatransferred from the monitoring device and measures thepatientrsquos activity information through the step number detec-tion algorithm and creates a profile about the patientrsquos healthinformation together with the amount of light illuminationto detect sun exposure

31 Development of the Watch-Type Monitoring Devices Inthe monitoring device location-tracing functions using aGPS sensor can monitor the present location and migrationroute of the patient The ambient light sensor measures theamount of sunlight illumination exposed to the device andrecords it The 3-axis acceleration sensor records the valueof the 119909- 119910- and 119911-axis coordinate values in real time Theserver can get the number of patientrsquos steps through the stepdetection algorithm

The values of the sensors are obtained through the realtime transfer of the data through Transmission ControlProtocolInternet Protocol (TCPIP) communication on theCDMA network After connection to the server througha Short Message Service (SMS) such as Server Open SMSand Transmission Close SMS for transfers the values of thesensors exchange data with each other At this moment thetransfer of the data by contacting the server is scheduledaccording to the regular cycle defined by the user The servercan inform the care professional or patient by alarm in thecase of special events such as injection time and escape frompatientrsquos safety zone of patient

Themonitoring device is designed to be worn easily usingthe form factor of a wrist watch and because it is held inposition by a clamp it can prevent a patient from takingit off or losing it Thus if a demented patient experiencesemergency or wandering symptoms the problem can bequickly dealt with The internal block diagram of the watch-type monitoring device proposed in this paper is shown inFigure 1

32 Development of the Health Management Server Theserver system is composed of the receiver module forreceiving the transmitted data from the monitoring devicethe health management module analyzing data and thewebpages performing management functions and patientmonitoring as shown in Figure 2

First the receivermodulemanages thewatchrsquos connectionthrough the SMS receiver while waiting for the monitoringdevicersquos SMSThe receiver module with the Connection SMSreceives the accumulated data saved in themonitoring deviceas the defined protocols after assigning a socket and a threadusing TCPIP communication

The health management module generates the patientprofile by analyzing the transferred data It checks whetherthe user moves out of the scope of the designated safety zoneor not using the GPS sensor data And it converts the ambient

Journal of Applied Mathematics 3

SPK

CPUARM cortex-M3

Lightsensor

GPSmodule

LDO

Accelationsensor

OLED

Mini USB

Charger

CDMA moduleDTW600-W KT

Li polymerbattery

AMP

U-SIMANT ANT

Buttons times 4096998400998400128 times 64

Figure 1 Watch-type monitoring device and its internal block diagram

Receivermonitoring module

SMS receiver TCPIP server

Health information management server

Locationanalysis

Lightanalysis

Movementanalysis

Data analysis

Location Light Movement

Userlocationboundarydetect

Detectpercentageoflight

Read sensordata

Save data

Alert moduleSMS senderRequest

Loadalert data

Web server

Profile manager

User healthdata set

Loaduserdata

EHR serverat hospital

Profile each user

Accumulatedhealth data

Google map

User health dataLoaduserhealthdata

Check alarm

Healthinformation

manager

Locationmovement

light

Graph tool

Loca

tion

Mov

emen

t

Ligh

t

DB X Y Z

Figure 2 The System Operational Scenario

light sensor data into a percentage from 0 to 100 accountingfor the patientrsquos exposure time to sunlight Finally itmeasuresthe amount of a patientrsquos movement by counting walkingsteps based on the step detection algorithm using the 3-axisacceleration sensor dataThe patientrsquos data acquired from thismodule is separately saved into the database The data in thedatabase is used and recorded in the profile of each patientand can be monitored through the webpage

The webpage is used to monitor the tracing location andhealth information of the patient obtained from the DBFirst of all a care professional can set up a communicationperiod between the monitoring device and the server andthe scope of the safety zone through the settings The serverindicates whether the traced patientrsquos location is within thescope of the safety zone or not and his present location andthe scope of the safety zone would be marked in a circle on


Walking and count number

Save DBdatesignalcount

Preprocessing to signal

Detect peakdetect feature Compare result

Figure 3 Preprocessing of accelerometer data

the map The amount of sunlight indicates the exposure statehourly as the time-axis and exposure-axis through the graphThe activity mass also expresses the number of walk hourlythrough the graph The health information can preserve thepatientrsquos health and safety because it monitors the patientrsquosstate through an activity list by time order amount of sun-light and location of the patient measured during outdooractivities

4 Walking Step Detection Algorithm

In addition to the location-tracking service for dementiapatients the system provides accurate walking step detectionfor use in healthcare The step detection algorithm uses a3-axis accelerometer to accurately detect a patientrsquos steps andfurther analyzes his activities

41 Experimental Design The experiment done in this paperuses the watch-type monitoring device to compare the actualsteps counted in 30sim60 secondswith the value detected by theaccelerometer under the same conditions Eight people tookpart in this experiment creating 170 data of 3 types of stepsmdashfast steps normal steps and slow steps every day Each data iscategorized in the database by experiment date time and thenumber of steps Stored results are preprocessed into energyvalues for peak picking and analysis of distinctive features ofthe walk Analyzed features are used to distinguish the stepand nonstep activities and the measured number of steps isthen compared to the actual number of steps counted

42 Preprocessing Data Figure 3 shows preprocessing of theaccelerometer data Each acquired x- y- z-axis data are in8 byte double data types recorded 80 times per second Itmakes the calculation more efficient using the Signal VectorMagnitude (SVM) values than using 3 values simultaneouslyfor each calculation SVM in this experiment is expressed asthe following equation (see Figure 4)

SVM = radic1199092119894+ 1199102

119894+ 1199112

119894 (1)

The accelerometer records 80 times per second and evencatches subtle movements Therefore even if the patient isstanding still the accelerometer will be recording constantlychanging values These subtle noise signals could result inerrors when measuring the number of steps In this paperwe have used the Moving Average Filter (MAF) to filter out

these noises preventing errors The MAF has low pass filterproperties and it can be expressed as follows

119879 [119899] =

1

5

(SVM [119899 minus 2] + SVM [119899 minus 1] + SVM [119899]

+SVM [119899 + 1] + SVM [119899 + 2])

=

1

5

2

sum

119898=minus2

SVM [119899 minus 119898]

(2)

Here the value of 119899thMAF is denoted by 119879[119899] and SVM [119899minus1] means (119899 minus 1)th SVM Figure 5 shows the result of movingaverage filter

43 Step Detection Algorithm The step detection algorithmproposed in this paper finds the peaks from the preprocesseddata and then counts the number of peak values that are overthe threshold value which is calculated from the data

First to pick out the peaks we find the waversquos meangradient by computing the average of the gradient of twobundles of data intervals If this value is greater than thethreshold value it is considered the start of the peak andwhen the mean gradient becomes a negative value this pointis put into the peak point candidate It is expressed as follows

119866119899=

SVM119899+1minus SVM

119899

119879119899+1minus 119879119899

Average of 119866119899=

119866119899+ 119866119899+1

2

(3)

The peak candidate includes waveform errors or noiseerrors The following method is used to clear out the errorsand find the genuine peaks First we find the peak candidateswith a time interval of less than 03 seconds Collected dataare acceleration data for detecting the number of steps sothe movements must show regular intervals of high peak andlow peak Therefore peak candidates in the low period arenoise values from the wrong movement Then we store thecandidate with high SVM values as the actual peak and dropthe values considered as errors

Detected peak values are affected by the patientrsquos footstepsand the height of the swinging of arms so the values includeindividual differences However every waveform of walkinghas high amplitude followed by low amplitude Thereforewe use this feature to derive a threshold value with themean amplitude over 1 second and collect the peaks over thethreshold value Figure 6 shows the result from the detectionof step peaks

44 Results of Experiment The proposed algorithm is testedwith the watch-type monitoring device with an embedded


SVM

Raw data

2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

2

4

6

8

0

10

12

14

14

16

16

16

18

18

18 20 22 24 26 28 30

2 4 6 80 10 12 14 16 18 20 22 24 26 28 30

X

X

X

Y

Y

Y

Z

SVM

Figure 4 Preprocessing convert raw data to SVM value

2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30

Moving average filter

SVM MVF

X

g g

Y

Figure 5 Preprocessing moving average filter

2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30

X

Detect peak

SVM MVFSVM peakPeak average

Y

Figure 6 Result from the detection of step peaks

accelerometer using an 80Hz sample rate attached to exper-imentersrsquo wrists and tested on fast steps normal steps andslow steps

To measure the accuracy of the proposed algorithm wecompared the actual sum of steps and the detected sum ofsteps derived with the algorithm The results of this methodshowed 947 accuracy in total 93 in fast steps 967 innormal steps and 96 in slow steps

As the pace gets faster the gradient of SVM tends to growlarger and the phase interval narrows resulting in highererror rates However in cases of normal and slow steps inwhich the amplitude is gradual results have a higher rateof finding the peaks correctly showing a closer value to theactual number of steps Table 1 shows the analyzed data fromthe 8 people taking part in the experiment

5 Patient Profile Management System

The purpose of this paper is to monitor daily health infor-mation to manage the dosage adjustment and health care ofdementia patients Measures of the amount of outdoor actionand the resulting information on momentum can be healthinformationThe patient profile management system profilespatientrsquos daily information Patientrsquos daily information canbe generated and the disappearance of the patient canbe prevented through position information by integratingpatient data received via a smart watch In this paper afunction that analyzes patientrsquos momentum and integratesreceived data is included to implement such a system

The amount of exercise analysis calculates the numberof steps measured by the acceleration sensor as momentumaccording to the rules After the acceleration sensor datareceived from the smart watch is integrated with data about apatientrsquos sex age weight and height stored in the server theintegrated data generates momentum information

51 Amount of Exercise Analysis The step count obtainedthrough the step detection algorithm can be used as data thatmeasures momentum The patientrsquos data which is basicallystored in the server includes age height weight and personalinformation and this data is used as the standard for measur-ing a patientrsquos stride and momentum

Themotion characteristics such as stationariness walkingand running and information corresponding to movingdistance and exercise time are needed in order to calculate themomentum The moving distance can be measured throughthe GPS sensor but it is difficult to measure the exact movingdistance due to errors of the GPS sensor and the differencebetween indoors and outdoors Therefore the method thatmultiplies stride by the number of steps is used to calculatethe patientrsquos moving distance in this paper The stride can becalculated by subtracting 100 from an individualrsquos height andmomentum can be calculated as shown in below

Amount of exercise

= Amount of energy consumption (Kcalminlowastkg)

lowast Exercise per minute (min) lowastWeight (kg) (4)


100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

2013-02-14

Figure 7 Function for monitoring GPS amount of activity

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Light

2013-02-142013-02-14

2013-02-14

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Life information

Tic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Exercise 51 602 603 654 709 74 75 778 78 804 81 79 72 59 58 57 58 562 551 54

Light 21 202 203 323 354 500 733 775 783 762 76 77 772 56 28 272 31 322 326 292

Outdoor N N N N N N N NY Y Y Y Y Y Y Y Y Y YY

122

Figure 8 Patient profile system

Energy consumption varies with motion characteristicsand bottom surface Table 2 shows energy consumptionwhenwalking on the basic asphalt

The monitoring system developed in this paper canmonitor a patientrsquos momentum current position and theamount of light through a web page by using the GPSroute information during outdoor activities step detectionand momentum detection Figure 7 shows the functions ofmonitoring server

52 Create Patient Profiles The patientrsquos profile includesthe patientrsquos momentum amount of light and indoor andoutdoor detection information by GPS The patientrsquos datais obtained in every cycle and the patientrsquos momentum iscalculatedThe calculated result is integrated and then stored

Figure 8 shows the screen applying the patient profilemanagement systemdeveloped in this paperThemomentumobtained from the patient is divided into momentum whichis converted into a percentage andmomentum converted into


Table 1 Experimental results

Lab no58 71 72 83 99 110 112 150 Total Accuracy

Fast stepUC 117 111 111 109 116 118 117 105 904 9303R 138 120 119 117 121 123 120 109 967

Slow stepUC 33 35 40 32 39 31 32 35 277 9602R 33 36 41 33 44 31 34 36 288

Normal stepUC 71 77 71 72 68 66 66 68 559 9677R 77 75 66 72 75 61 73 78 577

Total mean () 18321740 9471UC user countmdashThe number of steps counted by the userR result of algorithmmdashThe number of steps counted by the proposed algorithm

Table 2 Amount of exercise on asphalt

1min 2min 3min 10min50Kg 4 8 12 12060Kg 38 96 144 14470Kg 56 112 168 16880Kg 64 128 192 19290Kg 72 144 216 216100Kg 80 16 23 240

calories After being integrated with light data the profile canbe developed of a patientrsquos daily life The patientrsquos profile isupdated daily And it stores the daily information andmovingroute measured for a day If the data is accumulated thedoctor can determine a more exact dosage and treatmentmethod through the patientrsquos daily life data

6 Conclusion

In this paper we developed an ubiquitous health manage-ment system for dementia patients following the concept ofIoT It is composed of a watch-type monitoring device andserver that not only monitors patientsrsquo locations but alsomanages patientsrsquo health by determining patientsrsquo activityaccording to the data derived with the step detection algo-rithm along with the ambient light sensor and accelerometerAccording to the results of the experiments normal stepshave 96 accuracy in detection and on average showed 94accuracy

Typical medical services for dementia focused mainlyon tracking the patientsrsquo location to prevent a patient fromgoing missing or getting lost The system developed in thispaper provides and monitors the health information of thepatients as well as location tracking Further research basedon this work could include a more comprehensive analysis ofa patientrsquos activities such as running or sitting and extensiveapplication of the IoT paradigm

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgment

This research is supported by Seoul RampBD Program(SS110008)

References

[1] M H Tabert X Liu R L Doty et al ldquoA 10-item smellidentification scale related to risk for Alzheimerrsquos diseaserdquoAnnals of Neurology vol 58 no 1 pp 155ndash160 2005

[2] M Brahami A Baghdad Atmani and A Matta ldquoDynamicknowledge mapping guided by data mining application onhealthcarerdquo Journal of Information Processing Systems vol 9 no1 pp 1ndash30 2013

[3] Company Keruve 2008 httpwwwkeruvecom[4] U-Safe Gangnam 2009 httpwwwgangnamgokr[5] KT I-Search 2009 httpwwwktcom[6] G E Mead W Morley P Campbell C A Greig M McMurdo

and D A Lawlor ldquoExercise for depressionrdquo The CochraneDatabase System Reviews vol 3 Article ID CD004366 2009

[7] H Y Moon S H Kim Y R Yang et al ldquoMacrophagemigration inhibitory factor mediates the antidepressant actionsof voluntary exerciserdquo Proceedings of the National Academy ofSciences of the United States of America vol 109 no 32 pp13094ndash13099 2012

[8] B Kim T Kim H-G Ko D Lee S J Hyun and I-YKo ldquoPersonal genie a distributed framework for spontaneousinteraction support with smart objects in a placerdquo inProceedingsof the 7th International Conference on Ubiquitous InformationManagement and Communication (ICUIMC rsquo13) Kota Kina-balu Malaysia January 2013

[9] R S H Istepanian S Hu N Y Philip and A Sungoor ldquoThepotential of Internet of m-health things ldquom-IoTrdquo for non-invasive glucose level sensingrdquo in Proceedings of the 33rd AnnualInternational Conference of the IEEE Engineering in Medicineand Biology Society (EMBS rsquo11) pp 5264ndash5266 Boston MassUSA September 2011


[10] C Doukas and I Maglogiannis ldquoBringing IoT and cloudcomputing towards pervasive healthcarerdquo in Proceedings of the6th International Conference on Innovative Mobile and InternetServices in Ubiquitous Computing (IMIS rsquo12) pp 922ndash926Palermo Italy 2012

[11] L Atzori A Iera and G Morabito ldquoThe internet of things asurveyrdquoComputer Networks vol 54 no 15 pp 2787ndash2805 2010

[12] J Yang ldquoToward physical activity diary motion recognitionusing simple acceleration features with mobile phonesrdquo in Pro-ceedings of the 1st ACM International Workshop on InteractiveMultimedia for Consumer Electronics (IMCE rsquo09) pp 1ndash10 NewYork NY USA October 2009

[13] L Bao and S S Intillem ldquoActivity recognition from user-annotated acceleration datardquo in Pervasive Computing vol 3001of Lecture Notes in Computer Science pp 1ndash17 Springer BerlinGermany 2004

[14] J Baek G Lee W Park and B J Yun ldquoAccelerometer signalprocessing for user activity detectionrdquo in Knowledge-BasedIntelligent Information and Engineering Systems vol 3215 ofLecture Notes in Computer Science pp 610ndash617 Springer BerlinGermany 2004

[15] N Ravi N Dandekar P Mysore and M L Littman ldquoActivityrecognition from accelerometer datardquo in Proceedings of the 20thNational Conference on Artificial Intelligence (AAAI rsquo05) vol 20pp 1541ndash1546 Pittsburgh Pa USA July 2005

[16] HW Yoo J W Suh E J Cha and H D Bae ldquoWalking numberdetection algorithm using a 3-axial accelerometer sensor andactivity monitoringrdquo Korea Contents Association Journal vol 8no 8 pp 253ndash260 2008

[17] S H Shin C G Park J W Kim H S Hong and J MLee ldquoAdaptive step length estimation algorithm using low-costMEMS inertial sensorsrdquo in Proceedings of the IEEE SensorsApplications Symposium (SAS rsquo07) pp 1ndash5 San Diego CalifUSA February 2007

[18] Y H Noh S Y Ye and D U Jeong ldquoSystem implementationand algorithm development for classification of the activitystates using 3 axial accelerometerrdquo Journal of the KoreanInstitute of Electrical and Electronic Material Engineers vol 24no 1 pp 81ndash88 2011

[19] Y Luo O Hoeber and Y Chen ldquoEnhancing Wi-Fi fingerprint-ing for indoor positioning using human-centric collaborativefeedbackrdquoHuman-centric Computing and Information Sciencesvol 3 article 2 pp 1ndash23 2013

[20] J Ahn and R Han ldquoAn indoor augmented-reality evacuationsystem for the Smartphone using personalized PedometryrdquoHuman-centric Computing and Information Sciences vol 2article 18 pp 1ndash23 2012

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MathematicsJournal of


Mathematical Problems in Engineering

Hindawi Publishing Corporationhttpwwwhindawicom

Differential EquationsInternational Journal of

Volume 2014

Applied MathematicsJournal of


Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of


Mathematical PhysicsAdvances in

Complex AnalysisJournal of


OptimizationJournal of


CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of


Operations ResearchAdvances in

Journal of


Function Spaces

Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of Mathematics and Mathematical Sciences


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Algebra

Discrete Dynamics in Nature and Society



Decision SciencesAdvances in

Discrete MathematicsJournal of


Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Stochastic AnalysisInternational Journal of


creation of a personal profile for patients and monitoringa patientrsquos location and measuring the amount of sunlightillumination and walking step count to use as medical dataThe system helps dementia patients avoid the risk of beingmissing or lost by wandering symptoms

2 Related Works

Recently the concept of Internet of Things (IoT) has beenapplied in ubiquitous healthcare systems and services [8ndash10] IoT is a novel paradigm of technologies that intercon-nect everyday objects with each other through the Internetexploiting multiple wireless communication interfaces andadvancements in computing devices [11] With the spread ofsmart phones and tablets loaded with various sensors such asGPS and accelerometers higher quality services are providedto the users by connection of the information on theWeb andreal world [12]

With the advent of IoT research on numerous medicalservices for patients has been performed [9 10] Researchon wireless networking technologies for developing a mobilehealthcare environment has been carried out and it leads intothe concept of mobile IoT (m-IoT) which is a new healthcareconnectivity paradigm that interconnects IP-based commu-nication technologies such as IPv6 over low power WPAN(6LoWPAN) with emerging 4G networks for future Internet-based healthcare services [9] Typically healthcare servicesare comprised of the sensors acquiring biosignals and theservers processing the huge amount of biodata generatedfrom the sensors Service platforms that interconnect cloudcomputing distributed processing and high speed dataprocessing systems following the concept of IoT are beingresearched for efficient healthcare services [10]

Studies on human movement detection and behavioralpatterns have been carried out in various ways for health-care services The motion recognition algorithm based ona motion-tree is developed using the acceleration featuresof a mobile phone [12] The motion detection algorithmis one of the basic methods for detecting the number ofwalking steps [13 14] Human movements are distinguishedby a pattern recognition algorithm and a way of extractingvarious motions are developed from basic motion patternsand feature vectors of humans This function reads normaland abnormalmovements for example sitting standing andfalling down as well as the number of steps [15ndash18]

Position tracking using GPS is one of the data for mea-suring the momentum as well as the current position of thepatient in a healthcare system Recently research on indoorposition tracking methods using Wi-Fi or other positioningschemes are being carried out because it is impossible to geta GPS signal indoors [19 20]

3 Development of a Ubiquitous HealthManagement System

The system consists of a watch-type monitoring deviceand server The monitoring device includes a GPS 3-axisaccelerometer and ambient light sensor It is worn on

the patientrsquos wrists and periodically transfers his activityinformation to the server derived from his location andamount of light illumination detecting sun exposure Thencare professionals and doctors can monitor the patientrsquoshealth condition through thewebpage delivered by the serverThe server identifies the location through the patientrsquos datatransferred from the monitoring device and measures thepatientrsquos activity information through the step number detec-tion algorithm and creates a profile about the patientrsquos healthinformation together with the amount of light illuminationto detect sun exposure

31 Development of the Watch-Type Monitoring Devices Inthe monitoring device location-tracing functions using aGPS sensor can monitor the present location and migrationroute of the patient The ambient light sensor measures theamount of sunlight illumination exposed to the device andrecords it The 3-axis acceleration sensor records the valueof the 119909- 119910- and 119911-axis coordinate values in real time Theserver can get the number of patientrsquos steps through the stepdetection algorithm

The values of the sensors are obtained through the realtime transfer of the data through Transmission ControlProtocolInternet Protocol (TCPIP) communication on theCDMA network After connection to the server througha Short Message Service (SMS) such as Server Open SMSand Transmission Close SMS for transfers the values of thesensors exchange data with each other At this moment thetransfer of the data by contacting the server is scheduledaccording to the regular cycle defined by the user The servercan inform the care professional or patient by alarm in thecase of special events such as injection time and escape frompatientrsquos safety zone of patient

Themonitoring device is designed to be worn easily usingthe form factor of a wrist watch and because it is held inposition by a clamp it can prevent a patient from takingit off or losing it Thus if a demented patient experiencesemergency or wandering symptoms the problem can bequickly dealt with The internal block diagram of the watch-type monitoring device proposed in this paper is shown inFigure 1

32 Development of the Health Management Server Theserver system is composed of the receiver module forreceiving the transmitted data from the monitoring devicethe health management module analyzing data and thewebpages performing management functions and patientmonitoring as shown in Figure 2

First the receivermodulemanages thewatchrsquos connectionthrough the SMS receiver while waiting for the monitoringdevicersquos SMSThe receiver module with the Connection SMSreceives the accumulated data saved in themonitoring deviceas the defined protocols after assigning a socket and a threadusing TCPIP communication

The health management module generates the patientprofile by analyzing the transferred data It checks whetherthe user moves out of the scope of the designated safety zoneor not using the GPS sensor data And it converts the ambient


SPK

CPUARM cortex-M3

Lightsensor

GPSmodule

LDO

Accelationsensor

OLED

Mini USB

Charger


Li polymerbattery

AMP

U-SIMANT ANT






Locationanalysis

Lightanalysis

Movementanalysis

Data analysis




Read sensordata

Save data


Loadalert data

Web server

Profile manager

User healthdata set

Loaduserdata


Profile each user


Google map


Check alarm

Healthinformation

manager

Locationmovement

light

Graph tool

Loca

tion

Mov

emen

t

Ligh

t

DB X Y Z















SVM = radic1199092119894+ 1199102

119894+ 1199112

119894 (1)



119879 [119899] =

1

5


+SVM [119899 + 1] + SVM [119899 + 2])

=

1

5

2

sum

119898=minus2

SVM [119899 minus 119898]

(2)




119866119899=


119899

119879119899+1minus 119879119899

Average of 119866119899=

119866119899+ 119866119899+1

2

(3)





SVM

Raw data

2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

2

4

6

8

0

10

12

14

14

16

16

16

18

18

18 20 22 24 26 28 30

2 4 6 80 10 12 14 16 18 20 22 24 26 28 30

X

X

X

Y

Y

Y

Z

SVM


2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30


SVM MVF

X

g g

Y


2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30

X

Detect peak


Y










Amount of exercise




100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

2013-02-14


100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Light

2013-02-142013-02-14

2013-02-14

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Life information

Tic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Exercise 51 602 603 654 709 74 75 778 78 804 81 79 72 59 58 57 58 562 551 54

Light 21 202 203 323 354 500 733 775 783 762 76 77 772 56 28 272 31 322 326 292


122









Fast stepUC 117 111 111 109 116 118 117 105 904 9303R 138 120 119 117 121 123 120 109 967

Slow stepUC 33 35 40 32 39 31 32 35 277 9602R 33 36 41 33 44 31 34 36 288

Normal stepUC 71 77 71 72 68 66 66 68 559 9677R 77 75 66 72 75 61 73 78 577





6 Conclusion





Acknowledgment


References



























Volume 2014




Journal of











Journal of


Function Spaces






Algebra










SPK

CPUARM cortex-M3

Lightsensor

GPSmodule

LDO

Accelationsensor

OLED

Mini USB

Charger


Li polymerbattery

AMP

U-SIMANT ANT






Locationanalysis

Lightanalysis

Movementanalysis

Data analysis




Read sensordata

Save data


Loadalert data

Web server

Profile manager

User healthdata set

Loaduserdata


Profile each user


Google map


Check alarm

Healthinformation

manager

Locationmovement

light

Graph tool

Loca

tion

Mov

emen

t

Ligh

t

DB X Y Z















SVM = radic1199092119894+ 1199102

119894+ 1199112

119894 (1)



119879 [119899] =

1

5


+SVM [119899 + 1] + SVM [119899 + 2])

=

1

5

2

sum

119898=minus2

SVM [119899 minus 119898]

(2)




119866119899=


119899

119879119899+1minus 119879119899

Average of 119866119899=

119866119899+ 119866119899+1

2

(3)





SVM

Raw data

2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

2

4

6

8

0

10

12

14

14

16

16

16

18

18

18 20 22 24 26 28 30

2 4 6 80 10 12 14 16 18 20 22 24 26 28 30

X

X

X

Y

Y

Y

Z

SVM


2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30


SVM MVF

X

g g

Y


2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30

X

Detect peak


Y










Amount of exercise




100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

2013-02-14


100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Light

2013-02-142013-02-14

2013-02-14

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Life information

Tic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Exercise 51 602 603 654 709 74 75 778 78 804 81 79 72 59 58 57 58 562 551 54

Light 21 202 203 323 354 500 733 775 783 762 76 77 772 56 28 272 31 322 326 292


122









Fast stepUC 117 111 111 109 116 118 117 105 904 9303R 138 120 119 117 121 123 120 109 967

Slow stepUC 33 35 40 32 39 31 32 35 277 9602R 33 36 41 33 44 31 34 36 288

Normal stepUC 71 77 71 72 68 66 66 68 559 9677R 77 75 66 72 75 61 73 78 577





6 Conclusion





Acknowledgment


References



























Volume 2014




Journal of











Journal of


Function Spaces






Algebra




















SVM = radic1199092119894+ 1199102

119894+ 1199112

119894 (1)



119879 [119899] =

1

5


+SVM [119899 + 1] + SVM [119899 + 2])

=

1

5

2

sum

119898=minus2

SVM [119899 minus 119898]

(2)




119866119899=


119899

119879119899+1minus 119879119899

Average of 119866119899=

119866119899+ 119866119899+1

2

(3)





SVM

Raw data

2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

2

4

6

8

0

10

12

14

14

16

16

16

18

18

18 20 22 24 26 28 30

2 4 6 80 10 12 14 16 18 20 22 24 26 28 30

X

X

X

Y

Y

Y

Z

SVM


2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30


SVM MVF

X

g g

Y


2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30

X

Detect peak


Y










Amount of exercise




100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

2013-02-14


100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Light

2013-02-142013-02-14

2013-02-14

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Life information

Tic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Exercise 51 602 603 654 709 74 75 778 78 804 81 79 72 59 58 57 58 562 551 54

Light 21 202 203 323 354 500 733 775 783 762 76 77 772 56 28 272 31 322 326 292


122









Fast stepUC 117 111 111 109 116 118 117 105 904 9303R 138 120 119 117 121 123 120 109 967

Slow stepUC 33 35 40 32 39 31 32 35 277 9602R 33 36 41 33 44 31 34 36 288

Normal stepUC 71 77 71 72 68 66 66 68 559 9677R 77 75 66 72 75 61 73 78 577





6 Conclusion





Acknowledgment


References



























Volume 2014




Journal of











Journal of


Function Spaces






Algebra










SVM

Raw data

2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

2

4

6

8

0

10

12

14

14

16

16

16

18

18

18 20 22 24 26 28 30

2 4 6 80 10 12 14 16 18 20 22 24 26 28 30

X

X

X

Y

Y

Y

Z

SVM


2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30


SVM MVF

X

g g

Y


2

2

4

4

6

6

8

8

0

0

10

10

12

12

14

14 16 18 20 22 24 26 28 30

X

Detect peak


Y










Amount of exercise




100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

2013-02-14


100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Light

2013-02-142013-02-14

2013-02-14

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Life information

Tic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Exercise 51 602 603 654 709 74 75 778 78 804 81 79 72 59 58 57 58 562 551 54

Light 21 202 203 323 354 500 733 775 783 762 76 77 772 56 28 272 31 322 326 292


122









Fast stepUC 117 111 111 109 116 118 117 105 904 9303R 138 120 119 117 121 123 120 109 967

Slow stepUC 33 35 40 32 39 31 32 35 277 9602R 33 36 41 33 44 31 34 36 288

Normal stepUC 71 77 71 72 68 66 66 68 559 9677R 77 75 66 72 75 61 73 78 577





6 Conclusion





Acknowledgment


References



























Volume 2014




Journal of











Journal of


Function Spaces






Algebra










100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

2013-02-14


100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Light

2013-02-142013-02-14

2013-02-14

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

100

80

60

40

20

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Life information

Tic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Exercise 51 602 603 654 709 74 75 778 78 804 81 79 72 59 58 57 58 562 551 54

Light 21 202 203 323 354 500 733 775 783 762 76 77 772 56 28 272 31 322 326 292


122









Fast stepUC 117 111 111 109 116 118 117 105 904 9303R 138 120 119 117 121 123 120 109 967

Slow stepUC 33 35 40 32 39 31 32 35 277 9602R 33 36 41 33 44 31 34 36 288

Normal stepUC 71 77 71 72 68 66 66 68 559 9677R 77 75 66 72 75 61 73 78 577





6 Conclusion





Acknowledgment


References



























Volume 2014




Journal of











Journal of


Function Spaces






Algebra












Fast stepUC 117 111 111 109 116 118 117 105 904 9303R 138 120 119 117 121 123 120 109 967

Slow stepUC 33 35 40 32 39 31 32 35 277 9602R 33 36 41 33 44 31 34 36 288

Normal stepUC 71 77 71 72 68 66 66 68 559 9677R 77 75 66 72 75 61 73 78 577





6 Conclusion





Acknowledgment


References



























Volume 2014




Journal of











Journal of


Function Spaces






Algebra




























Volume 2014




Journal of











Journal of


Function Spaces






Algebra
















Volume 2014




Journal of











Journal of


Function Spaces






Algebra









Research Article Ubiquitous Health Management System with …downloads.hindawi.com/journals/jam/2014/878741.pdf · 2019-07-31 · paper, we develop an ubiquitous health management

Documents