POINT OF CARE ANEMIA DEVICE€¦ · Pumping Data Collection ... There are many types of anemia, but the team’s focus is microcytic, normocytic, and macrocytic anemia. Microcytic

POINTOFCAREANEMIADEVICE

BME400FinalReportDecember11,2015

TeamMembers:CarlyHildebrandt–TeamLeader

KatieBarlow–CommunicatorJeffWu–BPAG

LizzieKrasteva–BSACMichalAdamski-BWIG

Advisor:Dr.ThomasYenClient:Dr.PhilipBain

2

AbstractAnemia affects twenty-five percent of the global population, being most prevalent in the

underdeveloped countries of Africa. Most types of anemia are preventable, but lack of funding andresources forcompleteblood testspreventsclinicians frommakingaproperdiagnosisandsuggestingappropriatetreatment.Thegoalistodevelopacost-efficient,portable,andaccuratedevicetocountredbloodcellsandmeasuretheirmeancorpuscularvolume.Apreviousteamdevelopedaproof-ofconceptusingamicrofluidicschannelandacomputerinterfacetomeasuretheresistancechangeofmicroparticlesmoving through themicrochannel.Therearemany improvements tobemadeon thedevice,and thefocuswillbeonpumpingandfilteringmethods.Eachdesignwasevaluatedusingadesignmatrix,andtheteam decided to move forward with the syringe pump and cell filters. Cell filters were testing andstatistical analysis showed significant differences before and after filteringmeaningmore thanwhitebloodcellswerelostintheprocess.Activeandpassivepumpingtechniqueswereevaluatedusingdilutedporcine blood samples to determine themicrochannel’s ability to count cells. The fluctuations of thechannelresistanceduringactivepumpingledtoinconclusiveresultsandwhilepassivepumpingshowedafewpeakscorrelatingtocells,notenoughwerecountedtomakeanyclaimsatthistime.

3

TableofContentsIntroduction.........................................................................................................................................4

Motivation................................................................................................................................................4ProblemStatement..................................................................................................................................4PastWork.................................................................................................................................................4CompetingDesigns...................................................................................................................................6

DesignAlternatives..............................................................................................................................7PumpingTechniques................................................................................................................................7FilteringTechniques.................................................................................................................................9

DesignEvaluation...............................................................................................................................12DesignMatrices......................................................................................................................................12ProposedFinalDesign............................................................................................................................14

Fabrication/DevelopmentProcess......................................................................................................15Materials................................................................................................................................................15Methods.................................................................................................................................................15FinalPrototype.......................................................................................................................................15Testing....................................................................................................................................................16

Results...............................................................................................................................................18Filtration.................................................................................................................................................18PumpingDataCollection........................................................................................................................19

Discussion..........................................................................................................................................21Ethics......................................................................................................................................................21NecessaryChanges.................................................................................................................................21SourcesofError......................................................................................................................................22

Conclusions........................................................................................................................................22Futurework............................................................................................................................................23

References.........................................................................................................................................25

Appendix...............................................................................................................................................iDataAnalysis...........................................................................................................................................iiiMaterials..................................................................................................................................................v

4

Introduction

MotivationAnemiaisadeficiencyofthehemoglobinintheblood,usuallycharacterizedbyabnormalsize,

shape, and reduced number of red blood cells and quantified by measuring hemoglobin and meancorpuscularvolume.Commonsymptomsofanemiaarefatigue,dizziness,rapidheartbeat,andshortnessof breath. [1] There are many types of anemia, but the team’s focus is microcytic, normocytic, andmacrocyticanemia.Microcyticischaracterizedbyabnormallysmallredbloodcells,usuallyintherangeof 4 to6µm indiameter and less than80 fL volume, and is causedby irondeficiency.Normocytic ischaracterized by having normal-sized red blood cells (6-8 µm, 80-100 fL) but is caused by decreasedproduction or increased destruction of red blood cells, such as hemolysis. Finally, macrocytic ischaracterizedbyabnormallylargeredbloodcells,usuallyintherangeof8-10µmindiameterandgreaterthan100fLvolume,anditiscausedbyvitaminB12deficiencyorhypothyroidism.[2]

Anemia is very prevalent in underdeveloped countries, due most commonly to malnutrition.Anemia affects roughly twenty-five percent of the global population, but the highest prevalence is inAfrica,makingupaboutsixtypercentofthosegloballyaffectedbyanemia.Mosttypesofanemiaarenotonlytreatablebutcanbeprevented.However,lackoffundingandresourcesforcompletebloodtestsinthese developing countries block the ability for clinicians to properly diagnose anemia and suggesttreatment. [3] The client, Dr. Philip Bain, volunteerswithGlobal Brigades, an international non-profitorganizationthathelpscommunitiesaroundtheworldmeettheirhealthcareandeconomicneeds.[4]After volunteering inGhana,Africa, hewanted to initiate thedevelopmentof an inexpensive anemiadetectiondevicetoallowclinicianstomakeaproperdiagnosisandgiveappropriatetreatmentoptions.

ProblemStatementAnemia affects many people worldwide and disproportionately affects those in developing

countriesduetoalackinmedicalinfrastructuretoproperlydiagnosetheblooddisorder.Aportable,easytouse,andcost-effectivedevice isneededtodiagnosethecondition inthesecountriesatthetimeofinitialmedicalcare.Anemiacanbediagnosedbyevaluatingredbloodcellsizeusingthemeancorpuscularvolume(MCV).Thegoal is to fabricateamicrofluidicdevice thateffectivelymeasures theMCVof redblood cells to determine if a patient has normocytic, macrocytic, or microcytic anemia with resultscomparabletocurrentcellcountingtechniques.

PastWorkLastyear,ateamdevelopedaproof-of-conceptdeviceforthisproject,consistingofa

microfluidicschannelandcomputerinterfaceusingLabVIEW.Theyusedapassivepumpingmethodinoneinlet-outletpairtotransport10µmmicroparticlesthroughamicrochannel,andthentheymeasuredtheresistancechangeacrosstheotherinlet-outletpair(Seefigures1,2).Eachpeakinfigure3representedamicroparticlemovingfromtheleftsideofthechanneltotheright,whiletheheightofthepeakdeterminedtheparticle’srelativesize.

5

Figure1:Channeldesign.Theinletandoutletrepresenttheflowofthemicroparticlesthroughthechannel,andE1andE2representtheelectrodes.[5]

Figure2:Circuitschematicoftheelectrodesthatmeasuretheresistancechangeacrossthemicrochannel.[5]

Figure3:Voltagevstimewithcoincidenteventofamicroparticlemovingthroughthemicrochannel.[5]

6

Thereisalotofworklefttobedoneforthedevice,however.Ameetingwiththepreviousteamwasarrangedwherepotentialimprovementsontheirprototypewerediscussed.Theymentionedthedifficultyofpassivepumping,asaddingtheproperamountofsolutiontoboththeinletandtheoutletforthepassivepumpingtoworkistedious.Inaddition,theyneverwereabletotestwholebloodforthedevice,andafilterisneededforthedevicetoremoveunwantedbloodcells,suchaswhitebloodcellsandplatelets.

CompetingDesignsInadditiontoresearchingthepastworkofthepreviousteams,wealsolookedintoother

competingdesignsalreadyavailable.ThetwodesignsclosesttowhatwearetryingtoaccomplisharetheCoultercounterandtheLabonaChipBloodCountingDevice.Bothdeviceswouldworkverywellintheoryforsolvingthecurrentproblem,butalsooffersignificantdisadvantagestothecurrentworkthathasbeendone.

CoulterCounterTheCoultercounteroffersmanyhospitalsandclinicstheabilitytocountandsizevariouscells

andparticlesinanelectrolytesolution.[6]Theadvantagesofthisdevicearethatwholebloodcanbeused,anditisversatileinthatitcanbeusedformorecountingstrategiesthansimplydetermininganemia.Despitebeingadvantageousinallowingmanycountingstrategies,thisisalsosomewhatofadisadvantagewhencomparedtoouridealdevicebecauseitisabitmorecomplexthanwhatweneed.Thedevicedoesalsocomeatasignificantcost-mostCoultercountersareontheorderof$2,000ormore,andwearelookingforadevicethatcanbeboughtinanunderdevelopedcountryforaround$50.Inaddition,theCoultercounterisabitlarge,thereforemakingitmoredifficulttotransport.

Figure4:Coultercounter[7]

LabonaChip-On-chipsamplepreparationforcompletebloodcountfromrawbloodTheseconddevicewefoundwascreatedattheUniversityofTorontoinToronto,ON,CA.The

devicecanalsousewholebloodandworksbyfilteringwhitebloodcellsfromredbloodcells,thendiluting,lysing,andmeasuringthecellsincountandsize.[8]ThedeviceismuchsmallerthantheCoultercounter,asitisamicrofluidicdevicemadeofPDMSandisthesizeofamicroscopeslide,makingthedeviceveryportable(seeFigure5).Thedevicealsousesapneumaticpressuredevicetodrivethe

7

fluidsthroughthevariouschannels,whichiseasiertousethanthepassivepumpingtechniquethepreviousteamusedforourdevice.However,thedeviceisstillmorecomplicatedthanwhatweneed,sinceweareonlyinterestedinredbloodcellcountandsize,andthepneumaticpressuredevicetodrivethefluidthroughiscostlyinitofitself,makingtheoverallcostofthedeviceontheorderof$500.Therefore,wedecidedtocontinueonwherethepreviousdesignteamleftoffonourdeviceandmakethenecessarychangesneeded.

Figure5:On-chipsamplepreparationforcompletebloodcountfromrawblood[8]

DesignAlternativesInordertocontinuewiththeprototypeanddesignofthepreviousteam,twocomponentswere

evaluatedforintegrationintothedesign.Whilethepreviousteamprovedtheeffectivenessofamicrofluidiccountingchannel,bloodwasneveranalyzedwiththedevice.Duetoblood’sheterogeneouscompositionandhighcellconcentration,itmustgothroughapreparationprocesspriortobeingsentthroughthechannel.Thisinvolvesfilteringthesample,soonlyredbloodcellscrosstheaperture,anddilutingtoalowconcentrationsothatonlyonecellpassesthroughthechannelatatime.Thetwodesigncomponentsthatwereevaluatedwerepumpingandfilteringtechniques.

PumpingTechniquesThepumpingmethodisnecessarytoevaluatethebesttechniquetopassthebloodsample

throughthedevicequicklyandeffectively.

PassivePumpThepassivepumpingtechniqueworksbyallowingthecellstotravelthroughthechannelsdown

theirconcentrationgradient.ThecountingchannelwouldbefilledwithPBSorsomesimilarconductingsolution.Whenadropletofadilutedbloodsolutionisplacedononeendofthechannelthesolutionwill

8

spreadandmovetotheoppositesideofthechannelsimplybecauseitwantstotravelfromhighconcentrationtolow.Anoverviewofthisprocesscanbeseeninfigure6.

Figure6:Overviewofpassivepumpingtechnique.Dropletofbloodsolutionisaddedtoinletendofchannelandbloodcellswillflowtowardstheoutletend,downtheirconcentrationgradient.

Thistechniquerequiresverylittleadditionalmaterials.Apipettewouldbeneededtoaddadropletofknownvolumetothechannelandthatisall.Thisallowsthisdesigntobeverylowcostperuse.Therearequiteafewdisadvantagesthataccompanythisdesign.Itisverydifficulttoplaceadropletonsuchasmallsurfaceandrequiresaskilleduser.Thisincreaseinuserdifficultycanhinderthetimeittakestorunthediagnostictestandincreasestheriskinwastingmaterialsiftheclinicianismorepronetomistakes.Thereisalsoverylittlecontrolovertherateofflowofthecellsthroughthechannel.Additionally,thiscanincreaseusagetimesanddecreasetheamountofteststhatcanbeperformedinsomeallottedamountoftime.Alsosincethereisnotanyactivepumpingtakingplace,thedirectionofflowmightnotbeunidirectionalwhichwouldgreatlyskewresults.

SyringePumpUsingasyringetomanuallypassthebloodsamplethroughthedeviceisanotherpumping

option.Thisdesignincludesasmallsyringefilledwiththedilutedbloodsolutionandatubingchannelthatconnectstothemicrofluidiccountingchannel.Theuserwouldmanuallycompressthesyringetosendthesampleslowlythroughthesample.Therateofflowcouldbemeasuredbasedonthespeedatwhichtheusercompressedthesyringeandthediameterofthesyringeopeningwherecellswillbeleavingandenteringthechannel.Theset-uptousingasyringeasthemodeofpumpingbloodthroughthechannelcanbeseeninfigure7.

9

Figure7:Syringeattachedtoamicrofluidicchannel.Userwouldmanuallycompresssyringetosendbloodsamplethroughthechannel.Thesyringewouldbeintegratedintothechannelthroughanadditionaltubeas

shown.

Someadvantagesofthisdesignistheoverallsimplicityandeaseoffabrication.Allcomponentsofthispumpcanbeboughtcommerciallyandeasilyintegratedintotheexistingdevice.Sincemanycliniciansarefamiliarwithoperatingasyringe,itshouldalsobesimpletouse.Thedisadvantagesincludeitstendencyforvariationbetweeneachtest.Becausetheflowratedependsmuchontheuserandthespeedatwhichtheusercompressesthesyringe,thismaycausedifferencesinflowbetweensamples.Theuserwouldneedtoapplyacontinuouspressureataspecifiedspeedtoperformthediagnostictestcorrectly.

PeristalticPumpAperistalticpumptosendthebloodsamplethroughthechannelisacommonmicrofluidic

techniqueandanotherdesignalternativethatwasconsidered.Thispumpincorporatesflexibletubingandarotortosendthebloodsolutionthrough.Therotorcompressesthetubingasitrotatespushingfluidthroughthetubingunevenly.Aperistalticpumpsimilartowhatwouldbeintegratedintoourdesignisshowninfigure8.

Figure8:Theperistalticpumptubingandrotorshownhere.Therotorrotates,compressingthetubingandsendingthefluid(whichcanbeseeninblue)throughthechannel.[9]

Theadvantagesofusingthistypeofmechanicalpumpisitsautomationandeaseofuse.Thepumponlyneedstheusertoloadthebloodsampleandthepumpwillautomaticallysendthefluidthroughthechannel.Becauseofthisautomation,theflowratewillalsobeconsistentthroughouteachtestandhavenodependencyontheuser.Becauseofthepumpsmechanicalpropertiesandvarietyofmovingparts,thisdesignwillrequiremoremaintenancethantheotheroptions.Foruseindevelopingcountrieswhereresourcesarelimited,thiscanbeveryproblematic.

FilteringTechniques Anemiacanbediagnosedbyanalyzingredbloodcellvolumeandsize.Bloodismadeupofmorecomponentsthanjustthedesiredredbloodcellsandinordertogetthemostaccuratemeasurement,afilteringtechniqueisnecessarytopreparethesamplesothatonlyredbloodcellsremain.Witha

10

filteringmechanism,thegoalistohaveonlyRBCssentthroughtheapertureofthemicrofluidiccountingchannel.

CellFilterAfterthecellsaredeliveredintothedeviceviaapumpingmechanism,theymustbefilteredto

isolatetheredbloodcellsfromtheothercomponentsofblood.Oneoptionisasyringefilter,showninFigure9,whichcontainsmixedcelluloseesterfiltermembranesinplastichousings.[10]Thisfilterwouldbeattachedtothepump,andcellsolutionwouldbepumpedthroughthefilterbeforeitissentthroughthemicrofluidicportionofthedeviceforanalysis.Thefilterporesizewouldbeabout11μminordertoexcludewhitebloodcells,whichhaveadiameterof12-15μm,andenableentryofredbloodcells,whichhaveadiameterbetween4-10μm.Thewiderangeofredbloodcellsizesreflectsthetypeofanemia,wheremicrocyticanemiaisdisplayedbyanRBCdiametercloserto4μm,andmacrocyticanemiaismanifestedbyadiametercloserto10μm.Thediameterofplateletsisontheorderof1-3μm,sotheywillpassthroughthefilter,buttheanalysisisnotpresumedtobeaffectedbytheirpresence.

Figure9:Asyringefilterisconnectedtothepumpandallowsforcellsolutiontobepumpedthroughthefilter,removingwhitebloodcells.[10]

Thissinglesyringefilterwouldbeeasytomanufacturesincethefilterunitsareavailableonthemarket,andtheywouldonlyneedtobeadaptedtothepumpandmicrofluidicchipportionsofthedesignusingasimplemethodsuchasattachingthepumpoutflowtothefilterusingatubeandthefilteroutflowtothechipusinganothertube.Operationtimewouldbelowwiththisdesignsincethecellsolutionwouldonlybepassingthroughasinglefilter.However,thisfilterwouldhaveshortlifespansinceitisnotreusable.Indevelopingcountrieswithfewresources,itmaybeunfeasibletoreplenishstocksoffiltersonaregularbasis.

CascadingFilter Anotherfilteroptionisamultiple-layeredcascadingfilterthatwouldbemadeupofthreefiltersofdecreasingporesize.Thefilterwouldbethesizeofa5mLsyringeandcontainthreedisksmadeofpolyethylenesecuredtothewallsofthecylindricalsyringe,asshowninfigure10.Thesethreedisksthatcontainperforationswillallowforsuccessivemovementofredbloodcellsthroughbasedondiameter.Atthetoplayer,allbutthelargestcells(WBCsof~15μmdiameter)willbeallowedtoflowthroughtothelayerunderneath,andthelayerunderneathwillcontainsmallerporesthatexcludecellsofa13μm

11

diameter.Thefinallayerwouldblockpassageofcellsunder11μmwide,effectivelyisolatingredbloodcellsfromwhite.

Figure10:Acascadingfilterisolatesredbloodcellsbyremovinglargerwhitebloodcellsfromthecellsolutionusingasuccessionofthreelayers,eachcontainingasmallerporesizethantheoneabove.

Thisdesignwoulddecreasecellaccumulationwithinthefilterbyhavingmultiplelayerswhichallowcellstoflowgraduallythroughthedeviceinsteadofbuildingupandclusteringtogetherabovethefilter.However,thedesignwouldbedifficulttofabricatesincethistypeoffilterisnotcurrentlyavailableonthemarketandwouldhavetobebuiltbythedesignteam.Inaddition,alongeroperationtimewouldbenecessarywiththisfiltersincethecellsolutionwouldhavetopassthroughmultiplefiltersinsteadofjustonefilterasinthecaseoftheaforementionedcellfilter.

Built-inMicrofluidicFilter Afinaloptionforthefilterisabuilt-inmicrofluidicfilterthatwouldbeincorporatedintothemicrofluidicchipapparatususedforcellanalysisbuiltbythepreviousdesignteam.Asshowninfigure11,arowofpillarswith11μmopeningswouldbeconstructedonthechiptoallowredbloodcellstopassthroughtheopeningswhileblockingpassageofthewhitebloodcells,whicharelarger.

12

Figure11:Abuilt-inmicrofluidicfilterconsistsofchannelsthatallowforpassageofredbloodcellsintothemicrofluidicdeviceandexclusionofwhitebloodcells.[11]

Thisbuilt-infilterwouldbeeasytousesinceitisintegratedintotheexistingdeviceandwouldallowfiltrationtobeaccomplishedwithoutusingaseparatephysicalcomponent.Ontheotherhand,thefilteritselfwillhavetobemadeonthemicroscale,suggestingthatfabricationwillbedifficultsincemuchsmallerpartsareinvolved.Additionally,whitebloodcellsanddebrismayclogthefilterquicklysincethereareonlyacertainnumberofopeningstoallowforcellstopassthrough.

DesignEvaluation Pumpingandfilteringtechniqueswereanalyzedandevaluatedbasedonavarietyofcategoriesinordertoproposethemosteffectivefinaldesign.

DesignMatrices

13

Figure12:Designmatrixforalternativepumpingtechniques.

Adesignmatrixwascreatedforthethreeaforementionedpumpingtechniques(Figure12).Thedesignswererankedonthesixfollowingcategoriesinorderofdecreasingweight:easeofuse,timeofuse,cost,easeofmanufacturing,size,andsafety.

Easeofuseisofhighprioritytoourclient,Dr.Bain.Hesaidthattheexperiencelevelofcliniciansatpointofcareclinicsindevelopingcountriesvaries.Hewantedourteamtodesignadevicethatwouldallowforcliniciansofanyskillleveltouse.Thepassivepumpingtechniqueisadifficultproceduretosetup,andclinicianswhohaveneverusedapassivepumpingtechniquebeforewouldneedtobetrainedtoperformthistechnique.Althoughaperistalticpumpisconvenientbecauseitcanregulatetheflowrate,thispieceofequipmentisexpensiveandbulky.Therefore,thesyringescoredhighestforeaseofuse.Cliniciansofvaryingskilllevelshavemostlikelyusedasyringebefore,eitherfordrawingbloodoradministeringshots.Theyarealreadyfamiliarwiththispieceofapparatus,sotheywouldnotneedanyadditionaltrainingtoutilizeourdevice.

Timeofuseisalsoofhighprioritytoourclientandwasweightedthesameaseaseofuse.Dr.Bainsaidpointofcareclinicsseeabout300-500patientsperday,soourdeviceneedstobequicktouse.Thesyringeandperistalticpumptiedinthiscategory,andthepassivepumpscoredthelowest.Thepreviousdesignteamsaidthatittookthemafewtriestogetthepassivepumptowork,sothatiswhyitscoredthelowestinthiscategory.

Thecostcategorywasdeterminedbyadditionalcoststothecurrentdevice.Thepassivepumpinginletsandoutletsarealreadybuiltintothecurrentdevice,sonochangesneedtobemadetothecurrentdesign.Thesyringeandperistalticpumpingtechniqueswouldrequireadditionalequipmenttobepurchased.Thiscausedthesetwotechniquestoscorelowerinthiscategory,withtheperistalticpumpscoringlowestbecauseit’ssignificantlymoreexpensivethantheothertwodesignalternatives.

Easeofmanufacturingwasbasedonhowmuchadditionalworkwouldbeneededtoadaptthepumpingtechniquestothecurrentdesign.Again,sincepassivepumpingisalreadypartofthecurrentdesign,itscoredhighestinthiscategory.Thesyringescoredsecondhighestbecausetheinletsofthecurrentdesignwouldhavetobemodifiedtofitthesyringe.Theperistalticpumpscoredthelowestbecauseinadditiontomodifyingtheinletsofthecurrentdesign,theotherendoftheperistalticpumpwouldhavetobefittedtothebloodsamplecontainer.

Sizewasbasedoffofthecurrentdevice.Therefore,thepassivepumpscoredhighest.Thesyringescoredsecondhighestbecauseit’sanadditionalpieceofequipmentthatneedstobeintegratedintothecurrentdesign.Theperistalticpumpscoredlowestbecauseitisabulkypieceofequipmentthatismuchlargerthanthemicrofluidicdevice.

Safetywasnotaconcernforthethreedifferentdesigns.Clinicianswhoworkinclinicsaretrainedinhandlingbodilyfluids.Thethreeproposeddesignsdonotexposetheclinicianstoanythingtheyarenotalreadyexposedto.

14

Figure13.Designmatrixforalternativefilteringtechniques.

Aseconddesignmatrixwascreatedforthethreefilteringtechniques(Figure13).Thedesignswererankedonthesevenfollowingcategoriesinorderofdecreasingweight:easeofuse,timeofuse,cost,easeofmanufacturing,size,lifespan,andsafety.

Easeofuseandtimeofusefollowthesamecriteriaasmentionedinthepumpingtechniquessection.Thebuilt-infilterscoredhighestinthesetwosectionsbecausethefilteringcomponentisincorporatedintotheexistingdesignandnoadditionalequipmentisneeded.Thecellfilterscoredhighestforcostandeaseofmanufacturing.Cellfiltersarereadilyavailableandcanbepurchasedinbulkatalowcost.Althoughmicrofluidicdevicesarefairlyinexpensivetomanufacture,theresearchanddesignneededtogettoafunctionalprototypecanbequiteexpensive.Masksneededtofabricatethemicrofluidicsdevicescancostover$100permask,andit’snotguaranteedthatourmicrofluidicfilterdesignwouldwork.Thebuilt-infilterwillbeintegratedintotheexistingmicrofluidic,soitscoredhighestinthesizecategory.Thecascadingfilterandbuilt-infiltertiedforlifespanbecausethesetwodesignsarereusable.Safetywasalsonotaconcernforthesetechniquesasthecliniciansshouldbetrainedtoproperlyhandlebodilyfluids.

ProposedFinalDesign Asseenfromthedesignmatrices,thesyringeandcellfilterscoredthehighestandwillbeincorporatedintothefinaldesign.Thefamiliaritythatclinicianshavewithsyringeswasabigfactorinchoosingthisdesign.Inaddition,syringesarereadilyavailableandoflowcost.Similarly,cellfiltersarealsoreadilyavailableandoflowcost.Althoughcellfiltersarenotreusablelikethecascadingandbuilt-infilter,ourteamfeltthatdevelopingcountrieslackthenecessaryresourcesandtimetokeepthereusablefilterscleanedandmaintained.Thus,thecheaperandeasieralternativeistogowiththedisposablecellfilter.

15

Fabrication/DevelopmentProcess

MaterialsForthissemester’swork,filteringwasdeterminedtobeamajorareaoffocus.Forthispartof

thedesign,asamplepackofvariouslysized,roundfilterswaspurchased.Thesefiltershaveaporesizeofabout10microns;thiswasdeterminedtobeanadequatesizetoletredbloodcellsthroughwhileatthesametimetrappingthewhitebloodcellsinthecollectedsamples.

PDMSwasthematerialofchoiceforthefabricationofthemicrochannelsusedinthisproject.ThepreviousgroupthatworkedonthisprojectusedPDMSwithgeneralsuccessandasubstantialamountremainedfromtheirwork.Therefore,itwasusedagainforthecontinuationofthisproject.

SyringetochannelattachmentrequiredaseriesofLuerlocksandtubing.TheseweresecuredtothePDMSchannelinletportwithastrongclearadhesivetopreventleakageandbackflow.

MethodsThechannelitselfwasmoldedfromPDMS,vacuumedandallowedtosetoverthemasterdisc

designedbythepreviousteam.Itwasfabricatedsothatthethicknesswasapproximately1cmhigh.AfterthePDMShadcured,thechannelwasremovedfromthemasterdiscandplasmatreatedtoaglassslide.Thechanneldesigncontainedfourports,twoforelectrodeplacement,aninletandanoutlet.AllfourwerepunchedfromthePDMSusingabiopsypunch.ALuerlockadapterwasplacedintotheinletport,whereitwassecurelysealedwithastrongadhesive.Themicro-scalesizeoftheaperturerequiresastrongsealattheinletporttominimizeleakageandbackflow.Becauseofthis,thethicknessofthePDMSwasvitalinassistingthesecurementoftheLuerlockadaptersincethreadedendhasalargesurfaceareatoholdontowithoutblockingtheopeningofthechannel.ThetubingconnectedfromtheLuerlockadaptertoanotheradapterwherethesyringeisattached.

FinalPrototypeInfigure14,thechannelisattachedatitsinletporttothetubingandaseriesofLuerlock

adapters.Theadaptersallowthesyringe,tubingandchanneltobemanagedseparatelywhenneedbe.Eachisabletobedisconnectedforreplacementofindividualpartswhichisnecessaryifthedeviceistobeusedinpractice.

Figure14:PDMSchannelwithsyringeconnectedthroughtubingandLuerlockingadapters

16

Theoveralldevicesizeandportabilityrequirementsledtothefabricationoftheboxencompasseddeviceasseeninfigure15.Allcomponentswerecombinedintoanelectricalboxinordertopresentthedeviceinaclean,easytousemanner.Thisincludedthecircuitry,theNI-DAQwhichwasusedfordataacquisition,andthechannel.Asyringeconnectstoaninletportatthetopoftheboxwhichattachestothechannelthroughtubing.TheboxallowstheNI-DAQtobeconnectedtoacomputerwiththeappropriatesoftwareforcollectingdata.

Figure15:Finalboxenclosure.Syringeattachesattopofboxandsendsadilutedbloodsamplethroughtubingandthemicrochannel

Testing Thecellfilteringdesignaswellastheactivepumpingsyringedesignwerebothevaluated.Todeterminetheeffectivenessoftheproposedfiltrationmethod,cellcountsweremeasuredbeforeandafterbeingfiltered.Experimentsweresetuptoevaluatetheabilityofthemicrochanneltocountcellsastheycrosstheapertureafterbeingmanuallypassedthroughbysyringe.

FiltrationToobtainanaccurateRBCcountusingthechannel,afilterwasdevisedtoseparateWBCsfromRBCs.The10umporesizewaschosentoexcludeWBCs,whichhaveadiameterof12-15um,fromRBCs,whichare6-8umwide.[13]Theporesize,whichisintermediatebetweenthatofredandwhitebloodcells,allowsmacrocyticRBCsupto10umindiametertopassthroughthefilter.Thefiltrationdevicewasfabricatedbyplacingacircularpolyestermembranefilterwithadiameterof47mmandaporesizeof10umontotheopenendofa50mLcentrifugetube.Aroundcutwasmadeontheplasticcaptoallowfilteredbloodtoflowdirectlyintoatube,andthecapwasscrewedon,securingandstretchingoutthepolyestermembrane(Figure16).AsolutionofbloodinPBSdiluted1:5000(1uLofbloodin5mLPBS)waspouredintothetubecontainingthefilter,andcellcountsoftheunfilteredandfilteredsolutionswereobtainedviauseofahemocytometer.

17

Figure16:Thefabricatedfilteringmethodusedtogatherfiltrationdata

Itisexpectedthatcellcountbeforeandafterfilteringwillgiveanindicationoftheamountofcellslostthroughtheoverallfilteringprocess.Forthepurposeofthisdevice,ideallyonlylargewhitebloodcellswillberemovedthroughouttheprocess.Whitebloodcellsmakeuponly1%ofwholeblood,soinordertosupportthehypothesisthatonlyasmallportionofcells(whitebloodcells)arebeingfiltered,thecellcountsbeforeandafterfilteringneedtoshownosignificantdifference.[14]

PumpingandDataAcquisitionTrialswereperformedusingbothactiveandpassivepumpingtechniquestosendadiluted

bloodsamplethroughthedevice.Electrodeswerefittedeachtimetomeasurechangesinvoltageascellspassedthroughtheaperture.

ActiveSyringePumpingThechannelwasfittedwiththeLuerlockadaptersandtubingtotesthowwellthechannel

couldmeasurechangesinvoltageascellspassedthroughitsaperture.Thechannelwasplacedunderamicroscopesothemovementofcellscouldbeviewedandmatchedwithcollecteddata.Avoltagedividercircuitwascreatedwiththechannelrepresentingoneresistorofthatcircuit.(FigureFROMPASTWORK)Ascellspassthroughtheaperture,theresistanceofthechannelisexpectedtoincrease,thereforeincreasingtheoutputvoltagewhichismeasuredthroughanNI-DAQdataacquisitiondevice.Theoverallset-upcanbeseeninfigure17.

18

Figure17:Theexperimentalset-upusedforactivefiltering.Themicrochannelisplacedunderthemicroscopewhileelectrodesfromavoltagedividerareattachedatportsinthechannel.AnNI-DAQ

measurestheoutputvoltage.

Aporcinebloodsamplewasdiluted1:5000inPBS.Thesyringewasfilledwiththedilutedbloodsolutionandattachedtothetubingandchannel.Labviewsoftwarewasrunasthebloodsamplewasmanuallypassedthroughthemicro-channel.Videowasalsorecordedascellsflowedthroughtheaperture.

PassivePumpingAnidenticalset-upasabovewasperformedwithoutthesyringeandtubingattachment.Afresh

PDMSchannelwasplasmacoatedtoaglassslideandplacedunderthemicroscope.Electrodeswereplacedintheportsofthechannelastheywerewithactivepumpingtrialsanddatawasrecordedascellspassedthroughtheaperture.

Withpassivepumping,adropofdilutedbloodsolutionwasplacedontheinletportofthechannelandcellsflowedfreelydownthechanneltowardstheoutletport.

Results

FiltrationStatisticalanalysisrevealedthattherewasasignificantdifference(pvalue<0.0001)between

cellconcentrationintheunfilteredandfilteredsamples(seeTable1).Thelargedisparitybetweencellcountsindicatesthatthefilterwasnotworkingasexpectedsinceitwashypothesizedthatfiltrationwouldonlyremoveasmallportionofcells.Thisanalysisleadstothebeliefthatredbloodcellsarealsobeingremovedinthefilteringprocess,whichwouldaffecttheaccuracyoftheoveralldevice.Fromthisexperimentaldataandanalysis,itwasdeterminedthatafilterwouldnotbenecessarysincewhitebloodcellsonlymakeupasmallpartoftotalcellcount,whichshouldnotdrasticallyaffectcountingwithinthe

19

channel.Additionally,byaddingafiltrationcomponent,designcomplexitywouldbeincreased,whichisnotidealforasystemusedinlowresourcesettings.

Table1:CellConcentrationinUnfilteredvs.FilteredBloodSamples

PumpingDataCollection DatawascollectedinLabViewandanalyzedwithapythonscriptwrittenbypreviousteammemberRussellLittle.Theanalysissoftwarefiltersandfitsthedatatoacurvelevelingvoltagemeasurementsandremovinginstrumentaldrift.

Active Themotionofmanuallypumpingabloodsolutionthroughthechannelshowedtogivegreatfluctuationsinvoltage.(Figure18)Ideallytherewouldbeaconstantbasevoltagewhichrepresentstheresistanceofthechannelwithnocellspresentandascellspassthroughtheaperture,distinctspikesinvoltagearerecorded.Overmanytrials,noneoftheseexpectedspikeswereabletobedistinguished;evenmicroscopingvideoofthechannelverifiedthatcellswerepassingthroughtheaperture.Theconstantfluctuatingchangeinvoltageovertimeisbelievedtobecausedbythevaryingpressureappliedtothesyringe.Asdifferentpressuresareappliedtothechannel,changesinresistanceandnoisecausethedatatofluctuate.Therearemanymorecontributingfactorsthatpreventedtherecordingofvoltagechangesandarediscussedfurtherinthesourcesoferror.

20

Figure18:Voltageovertimeasbloodcellswereactivelypumpedthroughmicro-channel.(a)Therawoutput

data(b)Filtereddata(c)Curvefitdatatolevelvoltage

PassiveUtilizingthepassivepumpingtechniqueresultedinafewtrialswheredistinctdipsinvoltage

wererecorded.(Figure19)Withthemicroscopeitwasverifiedthatthesedipsdidcorrelatewithcellspassingthroughtheaperturebutnotallcellscausedacorrespondingchangeinvoltage.Overtimeresultsbecamemoreinconclusivewithminimaldistinctchangesinvoltage.Manytrialswereperformedbutduetothelackofvoltagechanges,resultswereinconclusive.Potentialexplanationsforthelackinsubstantialdataisdiscussedfurtherinsourcesoferrorandfuturework.

Figure19:Voltageovertimeasbloodcellswerepassivelypumpedthroughmicro-channel(a)Therawoutputdata(b)Filtereddata(c)Curvefitdatatolevelvoltage

21

Discussion Forthefilteringtestingresults,itisapparentthatthereisasignificantdifferenceincellnumberbeforeandafterfiltering.Thisisconcerningbecauseitisobviousthatourfilterwasremovingredbloodcellsandwhitebloodcellsfromthesolution,whentheoriginalideaofthefilterwastoremoveonlythewhitebloodcells.Becausewhitebloodcellsonlymakeupapproximately1percentoftheblood,thereshouldnotbeasignificantdifferenceinthecellcounts,butthewhitebloodcellswouldstillhavebeenremovedfromthesolution.[14]Thoughwedeemedthefilterunnecessaryfortesting,thereisstillapotentialforwhitebloodcellstoenterthedevice.Duetotheirsize,theycouldblockthechannelandpreventanaccuratereading.Therefore,furtherresearchandbrainstormingwillneedtobedonefordetermininganoptimalmethodforremovingthewhitebloodcellspriortothebloodenteringthechannel.

Forthepumpingresults,itwasfoundthatthesyringedidnotdeliveraconsistentflowrate,thereforecausingerraticfluctuationsinthedatacomparedtothepassivepumpingtechnique.Itisthoughtthatanautomatedpumpingsystemcouldsolvethisproblemtodeliveraconsistentflowrate.Inthe“LabonaChip”paperfound,theteamofresearchersfromTorontousesapneumaticpressuredevicetoconsistentlymovethecellsthroughthechannels.[8]Thistechniquemaybesomethingtoconsider,andisdiscussedfurtherinfuturework.

EthicsTheassessmentofthisdevicereliesontheavailabilityofappropriatebloodsamples.This

projecttakesasmallbloodsampleandrunsitthroughamicrochanneltotestforthepresenceofanemiainthepatientandwhattypeofanemiaitis.Itisimportantthatallbloodsamplesusedinthetestingofthisdevicecomefromconsentingsourcesandallapprovalneededpriortousinghumansampleshasbeenmade.

Theporcinebloodthatwasusedforthetestingofthisprojectwashumanelygatheredanddonatedtobeusedbyalaboratoryoncampus.

NecessaryChangesTheresultsgatheredabovedemonstratetheneedtoredesignsomedevicecomponents.The

fluctuatingvoltageobservedthroughmanuallypumpingsolutionthroughthechannelshowstheneedforanautomatedpumpingmechanism.It’spredictedthatapplyingaconstantpressuretothesyringewouldeliminatethemajorityofthesefluctuationsandyieldbaselineresultssimilartothoseofpassivepumping.

Duetobudgetconstraints,thechanneldesignfromthepreviousteamwasusedfortesting.Thepreviousteamdesignedthechanneltoworkwellwhilepassivepumpingbutsomeminortweakstothechannelcouldmakeitmorereliablewithactivepumping.Asuggestedchanneldesignisshowninfigure20.Therewasatendencyforsolutiontotraveldowntheshortestandnearestpathsothenewdesignincorporatestwoequallengthoutletports(anelectrodewillbeplacedineither).Thiswilleliminatethefavorabilityofoneoutletovertheotherandallowcomparableflowtoeach.Inadditiontoanew

22

channel,buildingcupsaroundallchanneloutlets,allowingsolutiontopool,wouldpreventtheoverflowofsolution.

Figure20:Apotentialfuturechanneldesignwheretheoutletportsareofequallengththereforeflow

doesnotfavoroneovertheother.

SourcesofErrorTherearesomeknownvariablesthatmayhaveaffectedtheresultsofthissemester’swork.In

ordertobuildthePDMSchannel,themastermoldfromthepreviousteamwasused.Overtime,thephotoresistofthemasterstartedtowearandflakeoff.Thismadethefabricationofeachchannelmoredifficultandalteredtheoveralldimensionsandshapewhichmayhavehadaneffectonthetestingresults.

Avarietyofionsolutionswereusedtodilutebloodsamplesincluding1XPhosphateBufferSolution(PBS),anisotonicNaClsolutionandahypertonicNaClsolution.Theconductivityofeachsolutionmayhavehadamoresignificanteffectontheresultsthanexpected.Thiswasatradeoffbetweenusingamoreconductivesolutionthatwouldaffectcellsovertimeandalessconductivesolutionthatwouldhaveminimaleffectsonthecells.

Channelblockingwasarecurringconcernwiththechannelbeingthattheaperturewasonly15uminsize.Anysortofdebrisorlargersizedcellcouldclogtheapertureandpreventfutureparticleflow.Theclogwouldinturnaffecttheresistanceofthechannel.WithpassivepumpingaclogcouldeasilybereversedbylightlytappingthetopofthePDMSbutactivepumpingcausedmoresevereeffects.Thepressureofthesyringelodgesthedebrisfurtherintotheaperturemakingthenearlyimpossibletoreleaseandrenderingthechanneluseless.

Adifferentpotentialsourceoferrorintheresultswastheageoftheporcinebloodthatwasused.Thebloodwasacquiredfromalocallaboncampusanditwasstoredinafridgewithoutanyothermethodofpreservingit.Itislikelythatovertimethebloodcouldhavedeterioratedandformedclumps.Thiswouldthrowoffthefilteringdataasitwouldbehardtostudyaconsistentsampleoftheblood.

Conclusions Aportable,easytousedeviceisneededtodiagnoseanemiaindevelopingcountrieswheretypicaldiagnosticmethodsarenotavailable.ThisprojectintendedtoimproveuponaparticlecountingmicrofluidicdevicewiththeendgoalofdevelopingadevicethataccuratelymeasuresMCVofredbloodcells.Thefocusofthisprojectwasontwomodificationstothepreviousmicrofluidicdevice.The

23

microfluidicchannelwasfittedwithasyringepumpingmechanisminordertoquicklymeasureredbloodcellsizeandbloodfiltrationwasanalyzedwiththeintentionofremovingwhitebloodcellsfromadilutedbloodsolution.

Theresultsfromfilteringadilutedbloodsampleshowedasignificantdifferenceincellcountbeforeandafterfiltration.Thisledtoaconclusionthatfilteringthebloodsampleremovesmorethanjustlargewhitebloodcellssincethesemakesuchasmallportionofwholeblood.Theremainderofthedesignforwentthefilteringprocess.Activeandpassivepumpingofadilutedbloodsamplethroughthemicrochannelwasanalyzedtofindthatactivepumpingcauseslargefluctuationsinchannelresistancereducingtheabilitytodistinguishsmallvoltagechangescausedbycells.Passivepumpingshowedevidenceofvoltagechangesascellspassthroughthechannelaperturebuttrialsovertimeshowedinconsistentresultswhichpreventedanysignificantconclusions.

FutureworkThereisamultitudeofworkthatcanbedoneonthisproject,themostpressingbeingto

redesignthedevicetofunctionaswasoriginallyintended.Thismeanstakingabloodsamplefromapatient,directlyinsertingitintothedeviceandgettingausablereadingthatwoulddiagnoseanemiaaswellasthetypeofanemiapresent,withinaminute,thatwouldbeaccuratetoabout95percent.Atthiscurrenttime,itwasshownthatsyringepumpingbyhandisnotpracticalbecauseitcreatesunsteadyflowwhichcausesavaryingsignalthatisnotreadable.Somesortofautomationwouldneedtobeintroducedintothedevicetocreateasmoothandcontinuousflow,however,thisisdifficultconsideringthecostandsizeconstraintoftheproject.

Moreworkwouldneedtobedoneinordertoimprovethefunctionalityofthechannels,thisincludesmakingtheflowmorepredictablethroughthechannel.Anotherkeyimprovementthatneedstobemadeistheoverallproductionofthemicrochannels,currentlyittakesoveranhourandahalftocreateasinglechannel.Thisprocessneedstobemadequickerandmoreefficienttohelpmaketestingnewdesignsfaster,aswellasmakingiteasiertofabricatelargerquantitiesatatime.Alongwiththis,abettermethodtoattachthetubingtothechannelsisneeded.Currently,thesealbetweenthechannelandthetubingleaksandthisleakagecouldharmthecircuitryovertime.

Usingthemicrochannels,acurvewouldbemadethatrelatesvoltagestodifferentsizedmicroparticles.Microparticlesofvariousknownsizeswouldbeused,andbecausetheyarespherical,theirvolumecouldbesimplycalculatedusingtheirdiameters.Thiswouldthenbeusedtocorrelatevolumestovoltages.Usingthiscurve,itwouldbepossibletodeducethesizeofthecellsthatpassthroughthechannelbyrelatingthevoltagetotherelatedvolume.

Someotherworkthatcouldbedoneinthefutureincludescleaninguptheprototypetobeapresentable,marketableproduct,alsomakingitmorecompact.ThiswouldalsoincludeswitchingfromtheNI-DAQunittoamicrocontrollertosimplifythecircuitryaswellasremovingtheneedforalaptoporcomputer.Atthispointascreenwouldbeattachedtothedevicetogivethereadoutsoftheresultsofthetest.Itwouldalsobebeneficialtotheendconsumertomakethevariouspartsofthedeviceeasiertoinstallandreplace,includingthenecessarymicrochannelsandthevariouswiringandelectrodesassociatedwiththecircuitry.

24

EventuallyitwouldbecomenecessarytogainIRBapprovaltotestwithhumanblood.Thiswouldbeanecessarysteptobeabletoprovethatthedevicefunctionswithhumanblood.

Anecessarystepthatneedstobedoneinthefutureiscontinuingtestingofthedeviceandoptimizingthedesignbasedonnetheobservationsandresultsoftheseteststomakesurethatitfunctionsasintendedandthattheresultsoftheanemiatestsareaccurateandreproducible.Thedevicewillbeupdatedasnecessaryuntilthisisachieved.

Anothernecessarystepinthefuturewillbetotestthefunctionofthedevicewithanaveragepersonwhohashadnoexperiencewithmicrofluidicchannelsordevicessimilartothisone.Thistestwillhelpmakeitclearwhetherornotthedeviceneedstobesimplifiedinordertobeusedbyageneraltechnicianatahealthclinicinathirdworldcountry.Inconjunctionwiththistest,ageneralmanualorprocedureofoperationforthedevicewillneedtobewrittenup.Thismanualwillbegivenatthetimeoftheabovetest,andwillberevisedasneededtobeunderstandableandexecutablewithoutanyotherpriortraining.Basedontheseresults,thedesignofthedevicewillbeadjustedasneededuntilthisgoalismet

25

References[1]Mayoclinic.org,'AnemiaSymptoms-MayoClinic',2015.[Online].Available:http://www.mayoclinic.org/diseasesconditions/anemia/basics/symptoms/con-20026209.

[2]V.Planche,S.Georgin-Lavialle.“Etiologiesanddiagnosticwork-upofextrememacrocytosisdefinedbyanerythrocytemeancorpuscularvolumeover130fL:Astudyof109patients,”AmJ.Hemotol.,vol.89,no.6,pp.665-666,2014.

[3]Who.int,'WHO|Worldwideprevalenceonanemia1993-2005',2015.[Online].Available:http://www.who.int/vmnis/anaemia/prevalence/summary/anaemia_status_summary/en/.

[4]Globalbrigades.org,'GlobalBrigades|GlobalBrigades',2015.[Online].Available:https://www.globalbrigades.org/.

[5]Chiang,M.,Choi,U.,Enge,J.,Little,R.BME402.“DevelopmentofAnemiaDetectionDeviceforLow-ResourceSetting,”2015.

[6]W.R.Hogg,W.Coulter;Apparatusandmethodformeasuringadividingparticlesizeofaparticulatesystem;UnitedStatesPatent3557352

[7]BeckmanCoulter.Multisizer™3COULTERCOUNTER®,ItemNo:6605697.Beckmancoulter.com

[8]J.Nguyen,Y.Wei,Y.Zheng,C.WangandY.Sun,'On-chipsamplepreparationforcompletebloodcountfromrawblood',LabChip,vol.15,no.6,pp.1533-1544,2015.

[9]Micro-milling.Digitalimage.Z-microsystems-Mouldingof'Lab-On-A-Chip'UnderCleanRoomConditions.Z-micosystems,n.d.29Sept.2015.[Online].Available:http://www.medicaldevice-network.com/contractors/nanotechnology/z-microsystems/z-microsystems5.html.

[10]Sigma-AldrichCo.LLC."Millex®SyringeFilterUnits,Disposable,MixedEsters."PoreSize0.22μm,Diam.33mm.N.p.,n.d.[Online].Available:http://www.sigmaaldrich.com/catalog/product/sigma/f8148?lang=en&region=US

[11]Shields,C.,Reyes,C.,Lopez,G.“Microfluidiccellsorting:areviewoftheadvancesinseparationofcellsfromdebulkingtorarecellisolation,”LabonaChip,2015.

[12]Guckenberger,DeGroot,Wan,Beebe,&Young.“Micromilling:Amethodforultra-rapidprototypingofplasticmicrofluidicdevice,”LabonaChip,15(11),2364-2378,2015.

[13]ThroughtheMicroscope.WadsworthCenter.NewYorkStateDepartmentofHealth.http://www.wadsworth.org/chemheme/heme/microscope/rbc.htm

[14]BloodBasics.AmericanSocietyofHematology.http://www.hematology.org/Patients/Basics/

Appendix

ProductDesignSpecificationsClient:Dr.PhilipA.BainAdvisor:ProfessorThomasYenTeam:

CarlyHildebrandt [email protected] [email protected] [email protected] [email protected] [email protected]

LastUpdated:Decmeber9,2015FunctionAnemiaaffectsmanypeopleworldwideanddisproportionatelyaffectsthoseindevelopingcountriesduetoalackinmedicalinfrastructuretoproperlydiagnosetheblooddisorder.Aportable,easytouse,andcost-effectivedeviceisneededtodiagnosetheconditioninthesecountriesatthetimeofinitialmedicalcare.Anemiacanbediagnosedbyevaluatingredbloodcellsizeusingthemeancorpuscularvolume(MCV).ThegoalistofabricateamicrofluidicdevicethateffectivelymeasurestheMCVofredbloodcellstodetermineifapatienthasnormocytic,macrocytic,ormicrocyticanemiawithresultscomparabletocurrentcellcountingtechniques.Clientrequirements

● Deviceshouldprovideanaccuratediagnosisofanemiaanddifferentiatebetweenmicrocytic,normocytic,andmacrocyticanemia

● Deviceshouldbelow-costandadaptabletoresource-limitedenvironments● Aclinicianshouldbeabletousethedeviceeasilyandreliablyafterpropertrainingwithan

intuitiveuserinterface● Thedeviceshouldbeabletodiagnoseanemiaatthepointofcare

DesignRequirements1. PhysicalandOperationalCharacteristics

a.Performancerequirements:Thedeviceshouldbeabletodiagnoseanemia,byidentifyingpatients’RedBloodCell(RBC)count,MeanCorpuscularVolume(MCV),andhemoglobin(Hb)levels.Testinganddiagnosistimeshouldtakelessthan30minutes.

b.Safety:Bloodsamplesmustbeintroducedtothedeviceinacontainedenvironmentsothatnocontaminationoccursbetweendeviceusesandbetweentheuser.Thedevicemustbeusedwithproperbloodcollectiontechniques.c.AccuracyandReliability:RBC,MCV,and/orHblevelsshouldbemeasuredwithatleast95%accuracywhencomparedtostandardcountingtechniques(i.e.CoulterCounter)toallowfordiagnosisofanemia.Diagnosiswith95%accuracyshouldtakenolongerthan30minutes.

ii

d.LifeinService:Devicehardwareshouldfunctionforatleast5years,andthebloodcollectionplatformwillbereusable.Softwareshouldbeabletobeupgradedwhennecessary.e.ShelfLife:Deviceandallattachmentsshouldhavetheabilitytobestoredfor5yearsfromthetimeofmanufacture.f.OperatingEnvironment:Thedevicewillbeprimarilyusedindevelopingcountries(i.e.countriesinAfrica,SoutheastAsiaandSouthAmerica),soavailableresourcesshouldbetakenintoconsideration.Generallyaccesstomostresourcesislimitedsothedeviceshouldbeabletostandaloneorrunwithminimaloutsidehelp(i.e.batteriesorasmallgenerator).g.Ergonomics:Thedeviceshouldbeeasilytransportableandusablebycliniciansofvaryingexperienceandeducationalbackgrounds.h.Size:Thedeviceshouldbesmallenoughtofitonabenchtopinaclinicalsettingwhilemaintainingenoughportabilitytotransporttoareasinneed.Thecircuitryhousingshouldbenolargerthan4”x6”x10”.Themicrofluidicdeviceshouldbenolargerthan4”x4”.i.Weight:Thedeviceshouldweighnomorethanfivepounds(2.3kg).j.Materials:Materialsshouldbelow-costanddurable.Theyshouldalsobebiocompatibleastonotdisruptcellstructureasthebloodismovingthroughthedevice(i.e.PDMSforthemicrochannels).k.Aesthetics,Appearance,andFinish:DeviceshouldhaveanoutputscreentoviewRBC,MCV,Hblevels,andclassificationofanemia.Datashouldbeeasilyinterpretedbytheuser.Circuitryshouldbehiddenfromview.

2.ProductionCharacteristics

a.Quantity:Oneprototypeisneededforproofofconcept.b.TargetProductCost:Productshouldnotcostmorethan$200.

3.Miscellaneous

a.StandardsandSpecifications:Tobedeterminedatalatertime.Devicewillbeusedindevelopingcountries,notcoveredbytheFDA.FDAguidelineswillbefollowed,butapprovalisnotneeded.b.Customer:Cliniciansofvaryingskilllevelsindevelopingcountries.

c.Patient-relatedconcerns:Shouldbeabletogivepatientadiagnosisatpointofcare.d.Competition:Existingdevicesincludethefollowing:

● CoulterCounter:AcoultercountermeasurestheMCV,hemoglobinandtheredbloodcelldistributionwidth.Thecoulterprinciple,whichusesthedirectcurrentimpedancemethod,governstheuseofthedevice.

● HemoGlobe:HemoGlobeisanon-invasivedevicedesignedbystudentsfromJohnsHopkinsUniversitythatmeasuresthehemoglobinlevelofabloodsample.Thedevicemeasurestheparameterandthensendsthedatatoacenterforfurtherdataanalysis.

iii

● MicrofluidiccardforRBCanalysis:Patent-US8034296B2:Thecartridgeenablesacompletebloodcountincludingredbloodcellanalysisofvariousparameters.Thecartridgechannelmayalsobesubjecttoanelectricormagneticfieldduringoperation.

DataAnalysisCreatedbyRussellLittleimportutilsfromparse_fileimportparsefromplotimportplotfromcalcimportpolyfit,find_peaks,mean,moving_averagedefprocess_raw_data(filename=None,maxima=None,binsize=20,deg=3,delta=.004,xlabel='Time(s)',title=None,**kwargs):'''Args:=======->maximaT/Fdeteckpeaksasmaximaorminima->labelThisfunctionisthemainentrypoint.givenafileitwilldothewholeprocessing...thing...thisismainlyjusttomakeiteasier.thisisverymuchdependantonthestructuerofthedatafile!Ifyouchangeacquisition!!!changethis!!!!'''#ASSUMPTIONS!!!!!!changethese.theremaybemoresubtleassumptions#furtherin.inputV='Voltage_0'#inputsignal,totrackphasechangesoutputV='Voltage_1'#outputsignal,totrackparticlesx='X_Value'#timearray,r=parse(filename=filename,**kwargs)#parsethedatafileassert(inputVinr)#__ASSUMPTIONS__weknowthistobetheinputsignalassert(outputVinr)#__ASSUMPTIONS__weknowthistobetheoutputsignal

iv

#addsomeusefulinfor['inputave']=[mean(r[inputV])]*len(r[inputV])r['outputave']=[mean(r[outputV])]*len(r[outputV])#============================='''Smooththeoutputtoremovelargepeaksresultingfromrarenoiseevents'''r['smoothedoutput']=moving_average(r[outputV],n=binsize)r['binsize']=binsize#============================='''removeinstrumentaldrift.Byfittingapolynomialtotheoutputsignal,thensubtractingit.'''r['polyfittooutput']=polyfit(r[x],r[outputV],d=deg)r['deg']=degr['outputadjusted']=r['smoothedoutput']-r['polyfittooutput']#=============================='''Runthedetectpeakalgorithmtocountthepeaksinthedatafile.'''mx,mn=find_peaks(r[x],r['outputadjusted'],delta=delta)ifmaxima:r['particles']=mxr['numberofparticles']=len(mx)else:r['particles']=mnr['numberofparticles']=len(mn)r['max']=maximar['delta']=delta#============================='''Thehardpartisdone.justplottheresults.Iwanttoproduceonesummarygraph.thatdisplaystheprocessingpipelineandonethatisjusttheresults.'''plots=[('RawOutput',r[outputV]),('SmoothedOutput',r['smoothedoutput']),('Result',r['outputadjusted'])]plot(r[x],plots,x_title=xlabel,plot_title=title,measurements=[])#=============================#andwe'redone.returntheresultsjustforgoodnesssake.returnr

v

process_raw_data()

Materials

Product PlacePurchased PricePolyester10umFilterSamplePack Steriltech 30.98ElectricalBox Amazon 7.45NI-DAQ BMEDepartment donated1,3mLSyringes BMEDepartment donatedLuerLocksandtubing Dr.ThomasYen donated