widesmartdrv

1TT:indicatesmodificationstopreviousversions-i-E.
D.
A.
EVENTDETECTIONANDANALYSISFORPHYSIOLOGICALSIGNALSManualofVersion1.
04Contents1LicenseAgreementiv1Basics1Overview1MainModulesandAdditionalRoutines1HardwareRequirements2SystemFiles3DataFormat4LogicofFileNames4TTDateFormat,VersionNumberandSupport52EventDetection61992Release20December2000E.
D.
A.
Manual1.
04-ii-InteractiveMode6CommonFields8UniqueFields11Graphics12Results13ParameterPre-definitions13TTBatchMode14TTMemoryResources14Methods15ElectrodermalActivity(EDA)15Electrocardiogram(ECG)18PowerSpectrum20Integration213EventCheck23GettingStarted23TTOn-lineHelpandInterruptKey23TTTheDefinitionFile24TheEntryScreen24TheUpperScreen24TTTheLowerScreen26TimeAnalysisandTimeFiles27TheGraphicInterface28ViewingtheDisplay28TTControllingtheAppearanceofaSingleChannel29ScrollingtheCurrentDisplay30PrintingaScreen30SelectionofResults31Measuring:InsertingandRevisingResults32TimeMarkings33ChangingGraphicColorSettings334ResultsandCalculations34TheResultFileConvertingUtility34ListingaResultFile34Options36Calculations36NumberofEvents36HeartPeriod,HeartRateandIBI36PhasicHeartRateChanges37TimeRelatedIBIStatistics38TTPlausibilityCheck39ResultFileStructure405ModifyingCODASDataFiles41E.
D.
A.
Manual1.
041992Release03/04/1993-iii-JürgenKayserHans-JürgenTrosienerPoschingerstr.
22Motzstr.
9D-12157BerlinD-10777BerlinTel.
+49(0)308559507+49(0)302169571+49(0)30314-25273Mail:kays0230@mailszrz.
zrz.
tu-berlin.
detrosienr@zrzsp3.
gp.
tu-berlin.
deReferences42Appendix43TTA.
TheDefinitionFileED.
DEF43TTB.
TheDefinitionFileEC.
DEF43C.
AnEDAResultFileListingExample45Index471992Release20December2000E.
D.
A.
Manual1.
04-iv-LICENSEAGREEMENTSOFTWARELICENSETHESOFTWAREANDTHEACCOMPANYINGWRITTENMATERIALSARETHESUBJECTOFCOPYRIGHT.
INCONSIDERATIONOFPAYMENTOFTHELICENSEFEETHEHOLDERSOFTHECOPYRIGHTTOTHESOFTWAREANDACCOMPANYINGWRITTENMATERIALGRANTYOUANON-EXCLUSIVERIGHTTOUSETHEPRODUCT.
THESOFTWAREMAYBERUNONASINGLEMACHINEATASINGLELOCATIONSOLONGASYOUCOMPLYWITHTHETERMSOFTHISLICENSE.
ANARCHIVALCOPYOFTHESOFTWAREISPERMITTED.
LIMITEDWARRANTYANDDISCLAIMEROFWARRANTYWARRANTEDISTHATTHEPRODUCTSHALLBEFREEFROMDEFECTSINTHEPHYSICALMEDIA.
THESOFTWAREANDALLACCOMPANYINGWRITTENMATERIALSAREPROVIDED'ASIS'WITHOUTWARRANTYOFANYKIND.
NOWARRANTYISGIVENTHATTHESOFTWAREISFREEFROMALLBUGSANDOMISSIONS.
FURTHER,ITISNOTWARRANTEDORGUARANTEEDANDNOREPRESENTATIONS,EXPRESSORIMPLIED,AREMADEREGARDINGTHEUSE,ORTHERESULTSOFUSE,OFTHESOFTWAREORWRITTENMATERIALSINTERMSOFCORRECTNESS,ACCURACY,RELIABILITY,CURRENTNESS,FITNESSFORAPARTICULARPURPOSE,OROTHERWISE.
THEENTIRERISKASTOTHERESULTSANDPERFORMANCEOFTHESOFTWAREISASSUMEDBYYOU.
E.
D.
A.
Manual1.
041992Release03/04/19932ThefirstversionofE.
D.
A.
waspresentedatthe20thannualmeetingoftheGermanPsychophy-siologicalSociety(DGPA)inHamburg(Trosiener&Kayser,1992).
3Martin&Venables(1980)isrecommendedasastandardhandbookfordescriptionandevaluationofmethodsinPsychophysiology.
11BasicsOverviewThesoftwarepackageE.
D.
A.
(EventDetectionandAnalysis)2analysesdigitizedphysiologicalrecordings3.
ItwasoriginallydesignedtoanalyzetheCODASdataformat(forCODASseechapter5,p.
41)butitisalsopossibletouseabinaryorASCIIformat.
Significanteventswillbedetected(e.
g.
,R-wavesofanECGorSCRsofanEDArecor-ding)andstoredinaresultfile.
Theseeventscanbedisplayedwiththeoriginalsignal.
Alterationsorcorrectionsoftheseeventsarepossible.
Thedetectedeventsorresultsarestoredinrelationtothebeginningofrecording,i.
e.
inrealtime.
Besidesalistingoftheresultsitisthereforepossibletolookatparticularpartsoftherecordinginordertocalculateseveraltonicorphasicparametersofinterestfromtheevents,e.
g.
SCRfrequencies,heartrate,heartperiod,phasicheartratesecond-by-second.
Ifoneoftherecordingchannelscontainstimemarkings,e.
g.
forstimulusonset,theprogramisabletodetecttheseanditispossibletoconsidertheeventsinrelationtothesetimemarkings.
Inshort,E.
D.
A.
enablestheusertogenerateparametersfromtherawsignal,thesortofdataneededforstatisticalpurposes.
MainModulesandAdditionalRoutinesThepackageincorporatesseveralindependentprograms.
First,thereistheroutineED.
EXE(EventDetection).
Tillnow,itispossibletoinvestigaterecordingsofelectro-cardiography(ECG)andelectrodermalactivity(EDA),tointegratethedatae.
g.
forelectromyography(EMG),andtorunapowerspectrumfordescriptivepurposes.
Theusermayspecifyseveralpresetspriortotheinvestigation,e.
g.
,typeofanalysis(ECG,EDA,integration,powerspectrum),timeconstant,amplitudecriterion,kindoffilter(high,low,band,notch,none).
ED.
EXEcanshowhowyourchoicesinfluencethedetectionofevents(e.
g.
,usingfilterswithdifferentupperorlowerbordersorusingdif-ferentamplitudecriteria).
Avalidatedsetup(thepresets)canbestoredinadefinitionfileED.
DEFtoruntheroutineinabatchjobforseveraldatafiles,e.
g.
overnight.
ED.
EXEalsotransformstheoriginaldata,i.
e.
youcanchoosebetweenabinaryoutputfilecontainingthefilteredsignal,theoriginalsignal(convertedfromASCIIorCODAS),orthesignalwiththetimeconstanteliminated.
AllfeaturesandtheusageoftheED1992Release20December2000E.
D.
A.
Manual1.
042routinearedocumentedinchapter2.
Second,theEC.
EXEroutine(EventCheck)displaysthedata(original,filtered,timeconstanteliminated)togetherwiththedetectedevents.
Uptoeightdatachannelscanbedisplayedatatimethereforemakingitpossible:toinspecttherecording;tocomparetheactivityinseveralchannelsbyoverlayingthedatacurves;tochangetheresolutionseparatelyforeachchannel;tojumptodifferenttimepointsintherecording;torevisethedetectedresults(change,add,or,deleteanevent),and,toselectortorejectthedetectedevents.
Further,youcaninvestigateaparticularchannelfortimemarkings,i.
e.
rectangularimpulses,whicharesimplesttheascentsordescentsor-morecomplicatedly-apatternoftworectangularimpulsesdeterminedbythetimeofandbetweentheimpulses.
Thedetecteddiscretetimepointsperformas'anchors'inordertodisplayarelatedtimeframeortimewindowonthescreen,e.
g.
tocollectinter-activelyonlythoseeventswhichfitinyourtimewindow.
TheECroutineaswellasthePSCREEN.
EXEroutinewhichprintsstoredgraphicscreenswhileworkingwithECaredocumentedindetailinchapter3.
TheRECAL.
EXEroutine(resultsandcalculations)convertstheresultfilestoASCIIcodeorcalculatesseveralparametersfromthedetectedand/orcorrectedevents,e.
g.
frequenciesforaparticularpartoftherecording.
Ofcourse,itispossibletodirecttheASCIIlistingtoanoutputfileandtocalculateownparametersfromtheresults,or-evenbetter-fromtheoriginalresultfilesforsuchacalculation.
Refertochapter4concerningtheRECALroutine.
TheCOSAM.
EXEroutine(CODASsampling)worksonCODASdatafiles,i.
e.
toextractpartsofadatafile,toreduceasamplingrate,tocombinetwodatafiles,orsimplytoinformyouaboutthenumberofchannels,samplerate,lengthofrecording,etc.
TheCOSAMroutineisdocumentedinchapter5.
HardwareRequirementsAllroutinesworkinaDOSenvironmentwithabasememoryof640KbRAMonanordinaryIBMPCoranycompatiblePC.
FortheEC.
EXEroutineacolormonitorwithaVGAorEGAinterfaceisrecommendedandisabsolutelynecessaryifyouaredisplayingmorethanonechannel.
FortheED.
EXEroutinesuchacolorscreenisrecommended.
Theotherroutineswillnotneedaspecialmonitortype.
Dependingontheamountofdataalarge,fastharddiskisrecommended.
ED.
EXEwilluseavailableextendedmemory(EMS)whichwillspeeduptheanalysis.
Allroutineswilloperatewithaninstalledco-processorwhichwillalsoincreasespeed.
So,bestisamachinewitha486CPU,built-inco-processor,16MBRAM,200MBharddiskandaVGAadapter-butthepackagewillworkwitha286-AT,noco-processor,E.
D.
A.
Manual1.
041992Release03/04/19933640KbRAM,20MBharddiskandaEGAadapter-onlymuchslower.
SystemFilesBeforecopyingthefilesyouareadvisedtomakeaback-upcopyoftheE.
D.
A.
distributiondisketteandtostoretheoriginaldatainasafeplaceasareference.
TheE.
D.
A.
distributiondiskettecontainsfourself-extractingarchivefileswhichstorethefollowingfiles(archivenamesintheleftcolumn):SystemFilesED!
.
EXEED.
EXE-toanalyzeyourrawdata(EDA,ECG,integration,powerspectrum)ED.
DEF-userdefineddefaultsettingsforED.
EXEED.
DBF/ED.
DBT-on-linehelpoftheentrymenuEC!
.
EXEEC.
EXE-tochecktherecordedandanalyzeddataEC.
DEF-userdefineddefaultsettingsforEC.
EXEEC.
DBF/ED.
DBT-on-linehelpoftheentryscreenEC.
HLP-on-linegraphiccommandkeyreferenceUTIL!
.
EXEPSCREEN.
EXE-toprintstoredscreensRECAL.
EXE-tolistandtocalculateonresultfilesCOSAM.
EXE-tomodifyCODASdatafilesRESULT.
PAS-resultfilestructure(TurboPascalcode)ExampleFilesDEMO!
.
EXEDEMO.
COD-CODASdatafileDEMO.
CAL-thecorrespondingcalibrationfileDEMO.
DEF-thecorrespondinguserdefaultsetup(cf.
EC.
DEF)DEMO.
D01-timeconstanteliminateddatachannel1DEMO.
D02-timeconstanteliminateddatachannel2DEMO.
G01-EDAresultsofchannel1DEMO.
G02-EDAresultsofchannel2DEMO.
G04-ECGresultsofchannel4DEMO.
ZMD-thecorrespondingtimefiletochannel5DEMO_EMG.
COD-CODASdatafilewithanEMGrecordingDEMO_EMG.
CAL-thecorrespondingcalibrationfileCopythesystemfilesafterextractionfromthearchivestoasubdirectory(e.
g.
,C:\EDA)ofyourharddisk.
Ifyouwouldliketotrytheexamplesmentionedinthismanualthencopythedemofilestoyourharddisk(e.
g,C:\EDAorC:\EDA\DEMO).
1992Release20December2000E.
D.
A.
Manual1.
044DataFormatED.
EXEandEC.
EXEworkonCODASdatafileswhicharestoredinan12-bitintegerformat.
Thedataisoptionallycalibratedtosignificantunits,e.
g.
,:mhoorMv,andthecalibrationconstantsarestoredinaCODAScalibrationfile(samefilenamewiththeextension*.
CAL).
Alternatively,binaryfilesmayserveasaninputtoED.
EXEandEC.
EXEconsistingofrealtypevalues(range2.
9x10-39to1.
7x1038,accuracy11-12digits,i.
e.
1signbit,39bitfractionand8bitexponent).
Thefirstdatapointholdsthesamplerate,thefollowingdatapointsrepresentthecontinuousdataflow.
Inaddition,ASCIIfilesmaybeusedasadatasource(onlyforED.
EXEwhichconvertsthesefilestothebinaryformat).
EachASCIIvaluemustbeseparatedbyacarriagereturnandalinefeed(ASCIIdecimalcodes13and10),i.
e.
eachlinecontainsonevalue.
Thefirstlinevalueconsistsofthetotalamountofdatapoints,thesecondvaluecontainsthesamplingrateinHertz.
ThefollowinglinesillustrateanASCIIfile.
Note,thatthereis1)noneedforanexponentialnotationand2)thecommentsintherightcolumnarenotallowedtooccurintheASCIIfile.
20001line1(20001datapoints)2.
0000000000E+02line2(frequency200Hz)-6.
3461538462E-02line3(firstdatapointatrealtime0sec)-6.
3781324248E-02line4(seconddatapointat0.
005sec)-6.
6665196610E-02.
.
.
-1.
2444866899E-01.
-1.
2791017588E-01line20003(lastdatapointat100.
000sec)LogicofFileNamesAlloutputgeneratedbythemodulesisnamedasfollows:.
isastringof1to8characters,aone-characterstring,andatwo-characterstring.
codesthenumberofthechanneltowhichthefilerefers,e.
g.
01forchannelone.
referstothetypeofdataortothekindofresultthefileexistsof.
Table1liststhepossibilitiesof.
E.
D.
A.
Manual1.
041992Release03/04/19935Table1.
E.
D.
A.
filenames*.
Bbinarydata:convertedASCIIdata(ED)*.
Cresultfile:controldata(ED)*.
Dbinarydata:filteredortimeconstanteliminateddata(ED)*.
FASCIIoverflowlisting(ED)*.
Gresultfile:detectedevents(ED)*.
Kresultfile:correctedevents(EC)*.
Rresultfile:realtimerelatedphasicheartrates(RECAL)*.
Zresultfile:correctedeventsincludingtimemarkings(EC)*.
ZMDtimefile(EC)Remarks.
Inbracketstheroutinewhichestablishestherespectivefiles.
*:filename,:channelnumber.
DateFormat,VersionNumberandSupportAlldatesinthismanualandinthesoftwareareenteredmm/dd/yyyy(month/day/year).
Incaseofinquiriescommunicatewithus(seecontactaddressesonpageiii)andrefertotherespectiveroutine,itsversionnumberandreleasedate.
Forexample,thefollowinglogopartsoftheEC.
EXE(shownwithinentryscreen,cf.
p.
26)andED.
EXEroutinesshowtherespectiveversionnumbers(1.
04and0.
26)andreleasedates(Jan.
1st,1993andOct.
12th,1992).
64444446444444ECVersion1.
04Release01/07/1993E4444447444444755EventCheck5555Copyright(C)JürgenKayser199255C)0)0))))))C)0)0)DD5500:00:00step**EventDetection**Rel.
0.
2610/12/92555500:00:00total**(c)H.
J.
TrosienerBerlin**55944444444444444444444N444444444444444444444444944444444444444444444N4444444444444444444444444444444N444444444444444444444484444444N444444444444444444444481992Release20December2000E.
D.
A.
Manual1.
0462EventDetectionOneofyourfirststepsinanalysisisthedetectionofinterestingeventsinyourdata,e.
g.
theoccurrenceofR-wavesinECGormaximainEDA.
Oneofthebestwaystobecomefamiliarwithourprogramsistotestthemwithour'demo'datafiles.
InteractiveModeStarttheprogrambytyping"EDkey)theentryscreenappearsandEDpromptsforinformation.
Thekeyterminatestheentryprocessinthecurrentfield.
Throughouttheprog-|ramthekeywillgiveyouanon-linehelp.
Ifyoupresstheprogram|quitsassoonaspossible.
|Intheentryscreenyouarepromptedforthe-filenameofyourdatafile-datatype-desiredmethod.
6444444444444444444464444444444444444444F1=HelpCtrl-Break=Exit44444444444444444444444414:41:254747555555FileDEMO.
COD5555DatatypeCODwithf=200.
00[Hz]T=0:1[h:m]40.
01[s]5555MethodEDA55Fig.
1.
PartialEDentrymask.
FileNameandDataTypeRememberthatyoucanchoosedifferenttypesofrawdata.
Thedatafilescanbeoneofthreedifferenttypes:ASCII,BinaryorCodas.
Thefilenameextensioncandefinethetypeofdataifyouuse'ASC'forASCII-files,'BIN'forbinaryfilesand'COD'forCodas-files.
Ifoneoftheseextensionsisspecifiedthenthefield'Datatype'isautoma-ticallysettothattype.
Ifyourdatafilehasanotherextensionoriftheextensiondoesnotcorrespondtoyourtypeofdatathetypemustbespecifiedindividually.
IfyouhavemultiplechannelsrecordedandyourdatatypeisnotCODthenyoumustchooseonechannelforanalysisandprepareitasabinaryorASCIIfile.
E.
D.
A.
Manual1.
041992Release03/04/19937Examplesforvalidfilenames:CHANNEL1.
BIN,VALUES2.
ASC,EXPER3.
COD,CHANNEL4,VP024,DATA0701.
EXPLInourexamplespleaseuseDEMO.
CODasfilename.
Thisfilecontainsof5channels(cf.
table2).
Table2.
RecordingChannelsofDEMO.
CODS)QChannelRecordingUnitS)Q1electrodermalactivity(EDA)onlefthand[:S]2electrodermalactivity(EDA)onrighthand[:S]3horizontaleyemovements(EOG)[°]4heartactivity(ECG)[mV]5timemarkingsofstimulusexposure[Mark]S)QMethodEDAcalculatesthetime,theamplitudesandthehalf-valuetimeofEDAreactions.
ThetypeofEDA-reactionisdetermined.
ECG1calculatesthetimeoftheR-wave.
ThecalculatedvaluesrepresenttheabsolutemaximaoftheR-wavecomplex.
Undersomecircumstances(e.
g.
smallsamplerates)theabsolutemaximumdoesnotrepresentthebestestimationofthetimeoftheR-wave.
ECG2calculatesthetimeoftheR-wavesasthebestestimation,i.
e.
itrepre-sentsa'virtual'maximum.
Usethismethodonlyif'ECG1'-resultsareinsufficient.
Theprocedureisverytimeconsuminganditmightbeveryconfusingforpresentationpurposesifthecalculatedmaximumisnottheabsolutemaximum.
POWcalculatesthepowerspectrumoftheselecteddatachannel.
Thedataaredividedinsuccessivesegmentsof1024datapoints.
Thespectraarelinearaveragedandtheresultisonlypresentedvisually.
INTintegratestheselecteddatachannel.
Theintegrationtimeconstantcanbeselectedoverawiderange.
Thedetectionprocessisskipped.
Afteryouhavespecifiedthemainfilename,thedatatypeandthedesiredmethodyouarepromptedwiththequestion:OK(Y/N)Ifyoutype'N'youcanmodifyallofyourentries(youmustusethekeytoskipthroughthefields).
1992Release20December2000E.
D.
A.
Manual1.
0486444444444444444444464444444444444444444F1=HelpCtrl-Break=Exit44444444444444444444444414:41:254747555555FileDEMO.
COD5555DatatypeCODwithf=200.
00[Hz]T=0:1[h:m]40.
01[s]5555MethodEDA5555Channelnumber1of55555Rangefrom-5.
00to5.
00[::S]5555Timeconstant1.
00[s]BESSELLOW-filt.
6.
order(even)5555Filterf,flow2.
00[Hz]fhi[Hz]savedata:NONE5555Displacement43.
00ptst=0.
215[s]5555Threshold:maximum0.
025[::S]Spec.
ParamsN(Y/N)5555Threshold:overlayedmaximum0.
025[::S]5555Threshold:speed0.
000[::S/s]5555Threshold:acceleration0.
000[::S/s]55555555AutoscaleNylo-6.
000yhi6.
000Overflows5555Maximumnumberattime[s]Pointnumber5555Timecompressionfactor(1.
.
9):155K)0)0)MK)0)0)M5500:00:00step**EventDetection**Rel.
1.
0401/07/93555500:00:00total**(c)H.
J.
TrosienerBerlin**55944444444444444444444N4444444444444444444444444444444N44444444444444444444448944444444444444444444N4444444444444444444444444444444N44444444444444444444448Fig.
2.
CompleteEDentryscreenwithdemodata.
Dependingonyourmethodthelowerpartofthescreenisfilledwithdifferentfields.
Somefieldsareuniquetothemethod,mostarecommontoallmethods(cf.
fig.
2).
LForourdemo-examplechoosethe'EDA'-optionasmethodandafterconfirmationthescreenwillchangetothe'EDA'-screen.
Ifthedata-fileshavebeenopenedsuccessfully,thelowerpartofthescreenisfilledwiththeappropriateentry-fields.
Thesamplerateandtheover-allsampletimeofyourrecordingisdisplayed(cf.
topoffig.
2).
CommonFieldsChannelNumberChannelnumbermustalwaysbespecified.
IfyouhaveCODASfilesthenumberofdatachannelsisdisplayedandtherequesteddatachannelisselected.
Forallotherfilesthechannelnumberisforidentificationpurposes(results,binaryoutput).
Range(from,to)Sometimesasignalmaybecutbecauseamplifierswereoverranged.
Allpartsofdataoutsideaspecifiedrangeareinspectedforsignalcuts.
Cutisdefinedhereas3ormoresuccessivepointswithequalamplitude.
Allcutsarereportedinanerrorfile.
Thenumberofcutsisshowninthefield'Overflows'.
Thisisanexamplepartofanerrorfilelisting(rawdatafromDEMO_EMG.
COD):BeginWidthValueBeginWidthPointPoints[::S|mV][s][s]E.
D.
A.
Manual1.
041992Release03/04/19934Lineartransformationsoftimeconstantaffecteddataarediscussede.
g.
byElbert&Rockstroh(1980)andThom(1988).
5Seee.
g.
CookIII&Miller(1992)orTietze&Schenk(1978)fordigitalfiltering.
943303-8.
212.
160.
001549633-8.
212.
480.
00151037848.
175.
190.
0020122337-8.
216.
120.
00351224438.
176.
120.
0015TimeConstantThespecificationofthisfieldisnecessary.
The(RC-)timeconstant(tc)mustbe$0.
0.
Itisdefinedasthetimeasignalneedstodecreaseto37%ofitsstationaryvalue.
Atcof0.
0indicatesthatyoursignalisnotaffectedbyatc.
Atc>0.
0leadstoaneliminationoftheeffectofthespecifiedtc.
Thisisdonebycomputingthesignalasifitwasaffectedbyatcof100seconds.
ThisisnearlyequivalenttoDCamplification4.
Somemanufacturersofpsychophysiologicalequipmentusetheterm'timeconstant'inaveryconfusingandunusualmanner(e.
g.
ACtimeconstant).
Indoubtdeterminethe(RC-)timeconstantofyourequipmentbyempiricaltests.
FilterType5FiltersarepredefinedastheBesseltype,whicharerecommendedbyconstantgroupdelayandlowoverswing.
Conceptuallythefilterisdesignedas(recursive)IIR-filter.
Theprecisedeterminationofsignalsphaseshiftbybesselfiltersmakesthistypemostinterestinginevent-relatedreactiondesigns.
Furthermore,aBesselfilterhaslowsignaldistortion(passbandripple)butlessselectivity(nosharpcutoffs).
FilterCharacteristicYoucanchoosehigh-pass,low-pass,band-passandband-rejection(notch)filters.
LOWLow-passfiltersareusedtoeliminatehighfrequencieslikeEMGinsomerecordings.
HIGHHigh-passfiltersareusedtoeliminatelowfrequencies,e.
g.
slowpotentialmovingartifacts.
BANDBand-passfiltersrejectlowandhighfrequenciesandallowonlyfrequen-ciesinthepassband(fromflotofhi)topasslikeEEG-filters.
NOTCHBandrejectionfilterscutoutafrequencyband(fromflotofhi)likelinenoise(50Hz,60Hz).
1992Release20December2000E.
D.
A.
Manual1.
0410FilterOrderThesharpnessofthecut-offsandthedistortionincreasewiththeorderofthefilter.
Ifyouneedaveryselectivefilterthenchooseahighorder.
Theorderofthefiltersareselectablebetween2and10forhigh-andlow-passfiltersinstepsof2.
Forbothbandfiltersthisisequivalentto1stto5thorder.
Cut-offFrequency(f,flo,fhi)High-andlow-passfiltersneedthecut-offfrequencyasacharacteristicparameter.
Cut-offfrequencyisthevaluewherethesignalgainis'3Dbdown'.
Fromthisvaluetheasymptoticgradientofthefrequencycharacteristichasavalueof-nx20dBperdecadewithn=orderofthefilter.
Bandfiltersalsoneedtheuppercut-offfrequency.
BinaryOutput(savedata)Youhavedifferentoptionstosaveyourdatainbinaryformat:ORIGTheoriginaldataaresaved.
ThisisnormallydoneifyouhaveASCII-data.
TCORThetimeconstantiseliminatedfromyouroriginaldataFILOYousavethefilteredsignalandifrequestedtheadjustedtimeconstant,too.
Thegroupdelayofthefilterisnotadjusted.
FILCYousavethefilteredsignalandalso,ifrequested,theadjustedtimecon-stant.
Thedataareadjustedinrespecttothegroupdelayofthefilter.
Thisisdonebycuttingthefirstpointsinthelengthofthegroupdelay(shift)andaddingvaluesattheendofthedata.
Thelastvalueisrepeateduntiltheoriginalsamplingtimeisreached.
NONENodataaresaved.
Intheinteractivemode,existingbinarydatafileswiththesamenamearenotdestroy-ed.
Amessageappearsonthescreenthatpromptsyoutorefuseoracceptoverwritingofbinarydata.
Inbatchmodedatafilesareoverwrittenunconditionally.
Displacement(groupdelay,shift)Thegroupdelayofafilteristhephaseshiftoftheincomingsignalwhichisproportionaltothesignalfrequency.
Besselfiltergroupdelayispracticallyfrequencyindependentanddeterminedbytheorderofthefilter.
Ahighorderproducesalargegroupdelayorshiftoftheincomingsignal.
Weestimatethegroupdelayandreportitindatapointsandseconds.
Ourestimateisthetimetheoutcomingsignalofthefiltersneedstoraisefrom0to50%ofthestationaryvalue.
WemustadjustourdetectedeventswiththisE.
D.
A.
Manual1.
041992Release03/04/199311value.
ThisisalreadydoneinallourED-resultfiles.
Ifyousavedataforfurtherusewiththe'FILO'-optionthebinaryoutputdataarenotadjusted!
UniqueFieldsSomefieldsarespecifictothemethod(s),someverysophisticated.
The'special'parameterscanimprovethedetectionofevents.
Threshold:MaximumAmaximumwithanamplitudebelowthisvalueisrejected,e.
g.
avalueof0.
5meansthatonlytheamplitudeofR-wavesequalorabovethisvaluearereported.
Threshold:OverlayedMaximum(special)TheamplitudeofanoverlayedEDAreactionmusthaveatleastthisvalue.
Defaultvalueisthemaximumthreshold.
Foranexplanationof'overlayedmaximum'refertofig.
5and6(p.
17).
Threshold:Speed/Acceleration(special)Speed(1stderivate)andacceleration(2ndderivate)controltheselectionofreactions.
Onlyreactionabovethepredefinedspeedandaccelerationareconsideredtobeofinterest.
Thevaluesarenormalizedtospeedandaccelerationpersecond.
Insomespecialcasesthedistinctionbetweenreactionandrecoveryshiftscanbeforcedbyvariationofaccelerationthreshold.
Thenonlyreactionswhichfittoacombinedcriterionofspeedandaccelerationarereported.
Normallytheparametersarenotused.
Window(powerspectrum)WindowingmeansaweightingdatatoreducespectralleakageerrorintheFastFourierTransform(FFT).
Thisbecomesnecessarysincethewaveformsaresampledforafinitelengthoftime.
Discontinuitiesattheendofthesegmentsproducespectralerrors.
Wehaveimplementedsomecommonwindows:RECTRectangularwindow.
Noweighting.
Leakageisverysevere.
HANNHanningwindow.
Thiswindowismostusefulwheregoodfrequencyresolu-tionisneededandamplitudeaccuracyisnotimportant.
HAMMHammingwindow.
Thiswindowisusefulforresolvingcloselyspacedfrequenciesandamplitudeaccuracyisnotimportant.
PARZParzenwindow.
Thedataareweightedwithatriangle.
1992Release20December2000E.
D.
A.
Manual1.
0412R-waveEnhancement(ECG2)ECG2convolvesthedatawithatemplateofaspecifiedfrequencyandenhancestheR-waves.
ByconvolutionwegetanenhancementoftherelevantpartsoftheR-wave.
UndersomecircumstancesthisleadstoabetterrecognitionofthemaximabutthereportedtimevalueswillnotcorrespondnecessarilytotheabsolutemaximumintheR-wavecomplex.
GraphicsResultsmaybedisplayedgraphically.
Asegmentof256or256xtimecompressionpoints(1.
.
99)isused.
Timecompressionmeansthatmoredatapointsareusedforaclearerpresentation.
Youcanrequestgraphicsonlyininteractivemodebytyping'y'whengraphics(y/n)appearsonthescreen.
Thebinaryresultfilesproducedingraphicalmodeshouldnotbeusedfurther,becausetherecursivefilteralgorithmleadstosomeproblemsifthesegmentlengthsaresmall.
Werecommendchoosingconvenientparametersandcheckingthisgraphically.
Youshouldthenrunthesamejobwithoutgraphicstoproducebestresults.
ScalingThegraphcanbescaledautomatically(Autoscale).
Yourdatasegmentisdisplayedsothatitfitsexactlyintothegraphicalsurface.
Thisisveryusefulifyouhavesomeunpredictablyhighsignalamplitudesorifyouwanttoinspectthevisualresultsexactly.
Manualscalingrequiresthespecificationofupper(yhi)andlower(ylo)borderofthescaling.
Dataoutofthatrangeareinspectedbutnotdisplayed.
E.
D.
A.
Manual1.
041992Release03/04/199313ResultsRangeOverflowsCuttingofyourdatabyoverrangedamplifiersmightoccur.
Thisisdetectedautomati-callyandreportedinanASCIIerrorfile.
Thereportgivesinformationaboutthebeginningandthelengthofthecutsignalinunitsofdatapointsandseconds.
Theamplitudeisreportedandthenumberofcutsisreportedinthefield'Overflows'oftheentrymask.
EventDetectionAlldetectedevents(SCRs,R-waves,.
.
.
)andrelatedvaluesarewrittentoaresultfile.
Tworesultfilesaregenerated.
One(main)holdstime,amplitude,typeetc.
ofevents,theotherholdscorrespondingspeedsandaccelerationstogiveideasforthesophisti-catedspecialparameters.
TheresultfilesrepresenttheinputforprogramEC.
OurEDAexampleproducesthemainbinaryresultfileDEMO.
G01.
ThismayislistedbytheprogramRECALwiththecommand"RECALDEMO.
G01SMARTDRV.
EXEorahardwarecache,be-causethisspeedsupprocessingextremelywhenusingvirtualstorageondisk.
7TheEDAsectionwasoriginallybasedonthealgorithmsdescribedbyThom(1988)butsub-stantiallyrevisedandextendedtoobtainamaximumofaccuracyandreliability.
15-Ifyouareintheinteractivemodeandnotusinggraphicdisplaymode(wheresomekindoffixedlengthsegmentsareused)youarepromptedforchangeofthesegmentlengthifEMSisnotsufficientforstorage(cf.
fig.
3).
-Inbatchmodeasegmentlengthparametercanbeset.
Ifitis0and$1000thenthisvalueisusedasthesegmentlength.
Avaluegreaterthanthemaximumnumberofdatapointsistruncated.
BesidesitsusefulnesstoforceEMSusage6,theconsequenceofanearlyprogramterminationproducedbyasmallamountofavailablediskspacecanbeoverriddenbyspecifyingasmallersegmentlength.
644444444644444444SegmentLength44444444744444444755Datafile20001datapts5555EMSallows11487datapts5555Werecommend20001datapts55555555Diskwillbeused.
Decrease#5555ofdatapointstouseEMS55555555Yourchoice20001datapts55555594444444444444444444444444444444489444444444444444444444444444444448Fig.
3.
SegmentlengthmaskMethodsElectrodermalActivity(EDA)7ThemeaningofallvaluesintheentrymaskisdescribedunderintheformersectionCommonFields(p.
8).
Asaresultofthedetectionprocessyoureceivethetimeofthemaximainrelationtothebeginningofyourmeasurement,theamplitudesofthemaxima,themaximatypesandthehalf-valuetimesifcomputingispossible(cf.
Results,p.
13,andAppendixC,p.
45).
1992Release20December2000E.
D.
A.
Manual1.
0416Fig.
4.
GraphicalpresentationoftheEDAmethod(resultscf.
AppendixC,p.
45).
Graphiccolorsaregivenintable3.
Fig.
4showsanormalEDArecording,treatedwithasixpole2Hzbesselfilterbeforethedetectionprocess.
Allmaximaandoverlayedmaximaexceedingaparticularamplitudecriterion(here$0.
025:S)aretreatedasmeaningful.
Inthegraphicalpresentationthey-axisscaleisfixedwithatimecompressionvalueof3.
Table3.
GraphicColorsforEDAmethodS)QcolorgraphicfeatureS)Qblack:(treated)datayellow:pointsthatareshiftedtothenextsegmentgreen:anglesidesofcomputedhalfrecoverytimeredplus(+):maximum,reactionwhiteplus(+):halfvaluetimewhitecross(x):computedhalfvaluetimelightblue:anglesidesofthemaximum,reactionS)QFivemaximaweredetectedinthisexample.
The1st,3rdand5th.
maximumareoftypeE.
D.
A.
Manual1.
041992Release03/04/199317Fig.
5.
SCRoftype2:Criticalestima-tionofhalfvaluetime(maximum4)(forcolorsseetable3).
Fig.
6.
SCRoftype2:Estimatedhalfrecoverytime(forcolorsseetable3).
1andthehalfvaluetimescouldbecomputeddirectly.
The2ndand4thmaximumareoftype2andthehalfvaluetimeshadtobeestimated.
Infig.
4youcanseethatformaxi-mum2thealgorithmleadstoavalidestimation.
Thepresentationoftheestimationofthe4thmaximumwassuppressedbecausethecomputedvalueisnotveryrealistic(cf.
fig.
5).
Anestimationisonlyrealisticifwecanextractagradientfromthedatawhichleadstoapracticaldeterminationofanapproximationwhichisneartotheassumedoriginalvalue(cf.
fig.
6).
Thisisonlypossibleifthereareenoughpointsforanes-timation.
Evenifthereareenoughpointsparticularcircumstancescanleadtoenor-mouserrors,e.
g.
acriticalsamplefrequency,acriticalAD-converterresolutionoraverylowamplitudecriterion.
Avisualinspectionoraprogrammaticalcontrolofplausib-ilityisstronglyrecommended!
ClassificationofMaximaAfterBoucsein(1988/1992)wedistinguishthreedifferenttypesofmaximaintheEDAmethod(cf.
table4).
Table4.
SCRClassificationS)QType1:Normalmaximum,halfvaluetimecouldbecomputed.
Type2:Overlayedmaximum,halfvaluetimehadtobeestimated.
Type0:Overlayedmaximum,halfvaluetimecouldnotbeestimatedbecausetheSCRhasnodistinctmaximum(see1streactioninfig.
7).
Boucsein(1988/1992)labelsthistypedifferentastype'3'.
S)Q1992Release20December2000E.
D.
A.
Manual1.
0418Fig.
7.
Reactionoftype0(1stSCR).
Thehalfvaluetimesarealwayscomputedifpossible,butnoplausibilitycheckismade.
Thegraphicalpresentationofthehalfvaluetimeissuppressedifthevalueisgreaterthantwicetherisetime.
Ahalfvaluetimeof-1iscodedwhenanestimationwasimpossible(normallytype0ofreaction).
Ahalfvaluetimeof-2iscodediftherearenotenoughpointsforanestimation.
Electrocardiogram(ECG)MostfieldsoftheECGscreenareknown(cf.
InteractiveMode,p.
6).
OnefieldisspecificformethodECG2:R-waveenhancement.
RememberthatyoucanchoosebetweentwodifferentmethodsforthedetectionofR-waves:ECG1andECG2.
ECG1determinestheabsolutemaximaintheneighbourhoodofalikelyR-wave.
Aconvolu-tionofthedatawithatemplateofspecifiedfrequencyisdoneinECG2.
Afrequencyvaluebetween8and10Hzisnormal.
YoucandeterminethefrequencywhenyoumeasurethewidthoftheR-wavewherefrequencyequalsthereciprocalofdoublewidth(inseconds).
ByconvolutionwegetanenhancementoftherelevantpartsoftheR-wave.
Undersomecircumstancesthisleadstoabetterrecognitionofthemaxima.
ThecomputedtimevaluesofECG2donotalwayspointtotherealmaxima(i.
e.
valuesfoundinyourdata)buttohypotheticaltimevalues.
Sometimesthisestimationisbetterthanthetimeoftheabsolutemaxima,primarilyifyouhaveusedaslowsamplerate,sothatthe'real'maximumliesbetweentwodatapoints.
MethodECG1isnormallyE.
D.
A.
Manual1.
041992Release03/04/199319Fig.
8.
ECGdemodata(channel4ofDEMO.
COD).
Graphiccolorsaregivenintable5.
sufficient.
Table5.
GraphicColorsforECGmethodS)QcolorgraphicfeatureS)Qblack:(treated)datayellow:pointsthatareshiftedtothenextsegmentredplus(+):Maximum,R-wavelightblue:anglesidesoftheR-waveS)QFig.
8showstheresultsofECG1.
Nofilteringwasdoneandallmaximaabove0.
4MvareselectedasR-waves.
Notimecompressionwasrequested.
AsaresultyougetthetimevaluesoftheR-wavesandtheiramplitudes:numbertimetypeamplitudesloperiserecovery[sec]maximumtime/210.
37591.
2390.
0000.
0000.
00021.
13091.
2920.
0000.
0000.
0001992Release20December2000E.
D.
A.
Manual1.
0420Fig.
9.
R-wavestogetherwithlargeT-waves(forcolorsseetable5).
31.
85591.
2910.
0000.
0000.
00042.
57591.
2160.
0000.
0000.
00053.
33091.
2230.
0000.
0000.
00064.
13091.
2170.
0000.
0000.
00074.
90091.
2610.
0000.
0000.
000Fig.
9showsanECG(notpartofthedemodata!
),wheresomeotherpartsoftheECGhavethesameamplitudeastheR-waves.
Theprogramdetectsthemaximawithoutfurtherintervention.
InbothmethodsECG1andECG2thetypeofmaximumissetto9.
Theslope-maxi-mum,risingandhalfrecoverytimearenotcomputedandsetto0.
0.
PowerSpectrumThepowerspectrumisusedtocontroltheeffectsoftherequestedfiltering.
Asintheothermethodsyouchooseadatachannelandspecifythefilter.
Awindowcanberequested.
WindowingmeansaweightingofyourdatatoreducespectralleakageerrorintheFastFourierTransform(FFT).
Normallythisisnotneeded.
E.
D.
A.
Manual1.
041992Release03/04/199321Fig.
10.
PowerspectrumofECG-chan-nel.
Fig.
11.
PowerspectrumofECG-chan-nelwithnotchfilter(48-52Hz).
Withthefields'flow'and'fhigh'youcanchangethex-scalingofthegraph.
Ifyouspecifye.
g.
40and60Hzyoureducethepresentationtotherangeof50Hznoise(withoutharmonics).
They-scalingisdoneautomatically.
Youcanchooselogarithmicorlinearscaling.
Themethoditselfdividesyourdatainsegmentsof1024pointsandcomputesthepowerspectrumofeachsegment.
Thesegmentsareaveragedwithoutanyweighting.
Theaveragedpowerspectrumremainsonyourscreen.
Fig.
10showsthepowerspectrumoftheECG-channel4withoutanytreatment.
Thescalingislogarithmicandarectangularwindowwasused.
Youcanseethatthesignalissuperposedbyalargeamountof50Hznoise.
Ifyouchoosea2ndorderbandrejectionfilter(notch)between48and52Hzthereductionof50Hznoiseisobvious(cf.
fig.
11).
IntegrationTheintegrationmethodintegratesthechosendatachannel.
Theminimumoftheusedintegrationtimeconstant(tc)hastobethesamplerate.
Ifyouusethissamplerateastcthesignalissimplyrectified.
Largertimeconstantsleadtoasignalintegration.
Thesignalamplitudeisweightedinrespecttothetimeconstant.
Noeventdetectionisdone.
Thegraphicalpresentationnormallyshowstheoriginaldatatogetherwiththeinte-grateddata.
Forclarificationthetwofiguresbelow(fig.
12and13)showthesignalsseparately.
Graphscalingisdoneinaccordingtotherangeoftheintegrateddata.
Ifyouusealargetctheoriginaldataaremostlyoutofrange.
1992Release20December2000E.
D.
A.
Manual1.
0422Fig.
12.
IntegratedEMG(forcolorsseetable6).
Fig.
13.
OriginalEMGdata(forcolorsseetable6).
Table6.
GraphicColorsforIntegrationmethodS)QcolorgraphicfeatureS)Qblack:(treated)datared:originaldataS)QFig.
12showstheintegrateddataofchannel1(EMG)withanintegrationtimeconstantof50ms.
Nofilteringandtc-eliminationwasused.
Thetimecompressionfactorforthegraphicdisplaywas16.
Fig.
13showsthecorrespondingoriginalEMGrecording.
LThedemofileDEMO_EMG.
CODholdstheoriginalEMGrecordingdata.
E.
D.
A.
Manual1.
041992Release03/04/1993233EventCheckTheEC.
EXEroutinedisplaysthedata(original,filtered,timeconstanteliminated)togetherwitheventsnormallydetectedbyED.
EXEandallowscheckingoftheevents,withoptionalmodificationorcorrection.
Thus,thisroutinemaycreatenewevent(result)files.
GettingStartedIfyoutype"EC/][[.
DEF]][]datafilenamefilenamefordefinitions(defaultextension.
DEF)options:/thishelp/bbinarydata/ccodasdataYoumayspecifyyourinputdatafileasthefirstcallparameterandthedatatypeeitherbytheoption(e.
g.
/bforbinarydata)orbytheextensionofthedatafile,e.
g.
*.
BINforbinarydata(seealsop.
6).
YoumaysetupyourparametersforECwiththehelpofanASCIItextfile.
YoutellECthenameofthisdefinitionfilewiththesecondcallparameter(cf.
chapterTheDefinitionFile,p.
24).
Ifthereisnosecondcallparameter,ECwillsearchperdefaultforafilenamedEC.
DEF.
Whencallingtheprogrambytyping"ECandkeysquittheentryfield.
On-lineHelpandInterruptKeyYouareallowedtopressthekeythroughouttheprogram.
Thiskeywillgiveyou|anon-linehelp.
Whileusingtheon-linehelpyoumayswitchtodifferenthelpmessages|bypressingor,thekeyquitsthehelp.
Youwillgetadifferent|setofhelpmessagesdependingonwhetheryouareworkingattheentryscreenorat|thegraphicdisplay.
|1992Release20December2000E.
D.
A.
Manual1.
0424The'emergency'interruptkeywillquittheprogramassoonaspossible.
|Openfileswillbeclosedandoccupiedmemorywillbeemptied.
Usethiskeyonlyifyou|areintrouble(!
)becauseincontrasttothenormalterminationofECwiththe|keynodataaresaved.
|TheDefinitionFileAdefinitionfileholdstheparticularparametersyouneedforEC.
EXE,e.
g.
significantlabelsforyourdata,thedisplayoroperationmode,orthekindofprinteryouuse.
NormallythisinformationisstoredinanASCIItextfilenamedEC.
DEFbutyoumayalsospecifyanotherfilenamewhencallingtheprogram(cf.
chapterGettingStarted,p.
23).
IfyouhavenodefinitionfileECwilltakedefaultsettings.
Youmayalterthedefaultsettingson-line.
Anexamplefile(EC.
DEF)isgiveninAppendixB(p.
43)toshowhowsuchadefinitionfilemightlooklike.
Allpossiblepresettingsarecommentedandarefurtherexplainedinthischapter.
IfyouhavecopiedthedemofilesDEMO.
*toyourharddiskyoucantrytorunECwith"EC.
EXEDEMO.
CODDEMO.
DEFhelp4444444444444444444444444444444447444444444444444444444444444444444755filenameDEMO.
CODtotaltime100.
000sec5555datatypeCODfrequency200.
00Hzsamplerate0.
005sec55K)))))0)))))0)0)0)0)0)0)0)0))))))))MK)))))0)))))0)0)0)0)0)0)0)0))))))))M55Ch.
**act.
**label**binaryfile****results****correction****select55551**Y**SCRleft**DEMO.
D01**Y**DEMO.
G01**Y**DEMO.
K01**Y**N55552**Y**SCRright**DEMO.
D02**Y**DEMO.
G02**Y**DEMO.
K02**Y**N55553**Y**EOG**DEMO.
D03**N**DEMO.
G03**N**DEMO.
K03**N**N55554**Y**ECG**DEMO.
D04**N**DEMO.
G04**Y**DEMO.
K04**Y**Y55555**Y**markings**DEMO.
D05**N**DEMO.
G05**N**DEMO.
K05**N**N55K)))))2)))))2)2)2)2)2)2)2)2))))))))MK)))))2)))))2)2)2)2)2)2)2)2))))))))M:555555Anymodificationsonentries(Y/N)N559444444444444444444444444444444444444444444444444444444444444444444444444444444894444444444444444444444444444444444444444444444444444444444444444444444444444448Fig.
14.
UpperpartofECentryscreenandcorrespondingprompt.
Column4tellsECanameofabinaryfile.
InthecaseofaCODASdatatypesuchaspecificationisoptionalbecausethemaindatafilecontainstheoriginaldataforallchannels.
Butyoumayspecifyabinaryfile,e.
g.
filtereddatabyED,todisplayotherdatathantheoriginalCODASdataforthischannel.
InthecaseofabinarydatatypeECexpectsatleastonebinarydatafiletobeactivated,i.
e.
thebinaryfilenameincolumn4mustbefollowedbya'Y'incolumn5(commonlythemainfilename).
IfECisunabletofindaspecifiedfilethecorrespondingcolumnissetto'N'(ECusesthecurrentDOSpath).
Column6(results)specifiesaresultfilewhichmaybeusedtodisplaydetectedeventsbyEDtogetherwiththeformerspecifiedsignal.
Again,ifECisunabletofindthefilethecorrespondingcolumnissetto'N'.
Column8(corrections)givesanametoanewoutputfile,i.
e.
aninteractivelycheckedand/orcorrectedresultfile.
Inthegraphicdisplayyoumayselectorde-selectaparticulareventfromtheinputresultfilefortheoutputcorrectionfile.
Further,youareabletochangeanevent,i.
e.
tochangethebasevalueorthemaximumofanevent,ortoinsertanewevent(cf.
chapterMeasuring,p.
32).
Ifyouwanttohavethepossibilitytocreateanewoutputresultfileyoumustspecify'Y'incolumn9.
BeforeoverwritinganexistingfileECwillaskforinformation.
ResultsfilesgeneratedbyECmayserveasinputresultfilesinanotherECsession.
Ifyouhavedecidedtogenerateacorrectionfilecolumn10(select)permitsyoutodeterminetoacceptaprioriallresultsfromtheinputfilefortheoutputfile(cf.
chapterSelectionofResults,p.
31).
Afteryouhavespecifiedthemainfilenameandthedatatypeyouarepromptedwiththequestion:"anymodificationsonentries(Y/N)".
Ifyoutype'Y'youcanmodifyanyoftheseentries(usethekeysandtochangethechannel,toskiptothenextentry,andtoreturntotheprompt).
Youmayrestore1992Release20December2000E.
D.
A.
Manual1.
0426theparametersettingsfromadefinitionfile(cf.
TheDefinitionFile,p.
24).
Aftertyping'N'allchannelswillbeinitialized.
DuetothearchitectureofDOSamaximumof15openfilesatatimeispossible.
Ifyouexceedthislimit,i.
e.
trytoopenmorethan15filesatatime,theinitializationisterminatedwiththemessage"toomanyopenfiles".
Notethatyouwillneedfilesinadditiontotheonesyouhavespecified:oneforeachcorrection,thetimefile,theCODASdatafile,andtheon-linehelpfile.
TheLowerScreenOnceyouhaveestablishedyourchannelentriesyouwilllookatamodifiedscreen(cf.
fig.
15).
555555compression:7currenttime:0.
000sec5555events:N555555555555555555555555changegraphendnewfile.
.
.
559444444444444444444444444444444444444444444444444444444444444444444444444444444894444444444444444444444444444444444444444444444444444444444444444444444444444448Fig.
15.
LowerpartofECentryscreenandmainmenu.
Bypressingyoucanadjustthesettingsofthelowerscreen.
Thecompressionratefixesthetimeresolutionofthegraphicdisplay,e.
g.
todisplayonlyevery7thdatapointasasinglepixel(cf.
fig.
15).
Acompressionof1isthenormalandlargesttimeresolution.
Currenttimerepresentsthetimeofthefirstdatapointattheleftborderofthegraphicdisplay.
Youmayspecifyanytimewithintherangeofthedatafile.
Eventssetsthemodeofcorrection,i.
e.
tousetimemarkingsasreferencesforcontrollingresultsofarecordingchannel(cf.
chapterTimeAnalysisandTimeFiles,p.
27).
ThekeywillforceECtoreadthedataandtoswitchtothegraphicdisplay(cf.
chapterTheGraphicInterface,p.
28).
Theandkeyswillsaveallcorrections(ifany).
Whiletheformerquitstheprogram,thelaterreturnstotheupperscreenandacceptsanewfilename.
Pressingwillshowversionnumber,releasedateandsupportaddress.
E.
D.
A.
Manual1.
041992Release03/04/199327TimeAnalysisandTimeFilesVeryoftenitisnecessarytolookatphysiologicalrecordingsinrelationtostimuluspresentation,e.
g.
specificSCRstoslides,or,phasicheartratechanges.
Ongoingevents(stimuli)arecodedsynchronicallyinformofrectanglesinanextrachannel.
ECcheckssuchatimemarkingchannelforrectangularimpulsesandstoresthetimesinatimefile.
Thesetimesserveas'anchors'forthe*.
Zresultoutputfileandforatimewindowdisplayinthegraph.
555555compression:7currenttime:0.
000sec5555events:Y5555timefile:DEMO.
ZMDtimewindowfrom:1.
000to:5.
000sec5555channel:5+)))),+)))),durationR1:0.
000sec5555height:1.
000)))-)))-R1.
))-.
))-R2.
))).
)))distanceR1/R2:0.
500sec5555edge:20123durationR2:1.
000sec55555555detectevents(timeanalysis)(Y/N)N559444444444444444444444444444444444444444444444444444444444444444444444444444444894444444444444444444444444444444444444444444444444444444444444444444444444444448Fig.
16.
Lowerpartofentryscreenforanevent(markings)relatedcorrectionmode.
Afterspecifying'Y'foreventsECpromptsyoutochooseanameforatimefile.
Thedefaultisthemainfilenamewithanextension*.
ZMD.
IfthereisanexistingfileECwillloadthefileandwilldisplaythetotalnumberofmarkingtimesandthefirsttimerefe-rence(cf.
fig.
16).
Thebordersofthetimewindowinthegraphicdisplayserveonlyasavisualhelp(twodottedlines).
Theyarerelatedtothecurrenttimereference.
Atimewindowborderofzeroindicatesvisuallythecurrentreferencetimepoint.
Ifthereisnotimefileyoumaycreateanewfile.
Forthispurposeyoumustspecifyanactivechannelwhichkeepstherectangularimpulses.
ECassumesabaselineofzeroforthischannelandsearchesforchangesfromthisbaselinewhichareatleastboftheheightoftherectangles.
Rectanglesofafixedsizeonlyrepresentabinaryinformation(onoroff),butthedurationofarectanglecangiveadditionalinformation.
ECmayanalyzeapatternoftwofollowingrectanglesandconsiderthedurationofbothrectanglesaswellasthedurationbetweentherectangularimpulses.
Asyouseeinfig.
16youmayspecifythispatternintheentryfieldsduration/distanceR1/R2.
Azerotimecodesanytime,i.
e.
ignorethispartofthepattern.
Forinstance,apatternof0.
000-0.
000-1.
000searchesfor1-secimpulses,apatternof2.
000-1.
000-0.
500searchesfora2-secimpulsesfollowedbya-secimpulsewithadelayofonesecond.
Thediscrete'anchor'isaedgewithinthispattern,codedbyanumberbetweenzeroandthree(cf.
schemaoffig.
16).
Ifallpatterntimesmatchzero,anevenedgenumber(0or2)forcesECtosearchallrisingedges,anoddnumber(1or3)tolookforallfallingedges.
Allofthisspe-cificationsmaybestoredinadvanceinadefinitionfile(cf.
chapterTheDefinitionFile,p.
24).
1992Release20December2000E.
D.
A.
Manual1.
0428Fig.
17.
ThewholeECgraphicdisplayshowing5activechannels(DEMO.
COD).
WhensearchingfortimemarkingsECcheckstimeswithaprecisionof±2xsamplerate(butatleast25ms).
Anongoingtimeanalysiscanbeinterruptedbythekey.
Ifyouthenconfirmtheinterruptionthetimefilewillcontainalltimepointsdis-coveredpriortointerruptionotherwisetheanalysiswillbecontinuedYoumayswitchbetweenreferencetimepointsinthegraphicdisplay(cf.
chapterTimeMarkings,p.
33).
Intheentryscreenyouarenolongerallowedtochooseancurrenttime(butyoucandosofromthegraphicdisplayusing).
AtimefilemaybeconvertedtoASCIIcodewiththeRECAL.
EXEroutine(cf.
chapterCalculationsandOtherOptions,p.
34).
TheGraphicInterfaceViewingtheDisplayThekeytogglesbetweentheentryscreenandthegraphicdisplay(cf.
fig.
17).
E.
D.
A.
Manual1.
041992Release03/04/199329Fig.
18.
Topleftofgraphicdisplay:filenameandcurrentlyactivatedchannel.
Fig.
19.
Bottomrightofgraphicdisplay:timeaxis,channelscaling,percentageofdataread.
WhengettingintothegraphicmodeforthefirsttimeorafterchangingtheactualtimeECwillreadthedata.
Inthegra-phicthekeyalwaysshowsanon-linehelp,theredrawsthescreen.
Themainpartofthescreendisplaysthesignalsoftherecordings.
Ifthereareresultswithinthetimerangeofthedisplaytheresultsareintegra-tedineachgraphwithortho-gonallinestogetherwiththelasttwodigitsoftheresultnumber.
Thetopleftcornerofthescreenexhibitsthemainfilename(cf.
fig.
18).
Belowthefilenameyouseethenumberofthecurrentlyactivatedchanneltogetherwiththecorrespondinglabel.
Resultparametersofthischannelarelistedbelow(number,timeofmaximum,amplitudeinunits).
Atthebottomofthescreenatimeaxisinformsyouaboutthetimescaling,startandendtimeofthecurrentdisplay,andtheper-centageofdatareadsofar(cf.
fig.
19).
Therightpaneldisplaysthescalingofeachchannelinrealunits.
Thenumberatthebottomofeachscalerepresentsanarbi-traryscalingfactorwhichmayrangefrom1to9999(seenextsection).
ControllingtheAppearanceofaSingleChannelIfyouwanttoselectanotherchannelinsteadoftheonecurrentlyactivatedyousimplyhavetotypethenumberofthischannel(1-8).
Onlythechanneldisplayedattheleftpanelmaybemodified.
1992Release20December2000E.
D.
A.
Manual1.
0430Thekeys,,,,andadjustthescalingfactorofthischannel.
Alternatively,youusetheorkeysandspecifythescalingfactordirectly(pressortochangetheformerscalingfactorandenterthenewone).
Thekeyrecalculatesthescalingfactorandthebestfittinghorizontalpositionofthegraphline(thisistheinitialdefaultdisplay).
Thekeyallowsyoutosetthescalingfactorofallmeasur-ingchannelsatonce,i.
e.
ofchannelscreatinganewcorrectionresultfile.
Perdefaultallchannelare'autoscaled'whichisexactlywhatthekeydoes.
Butitisalsopossibletopresetthescalingfactorforeachchannelviaadefinitionfile(cf.
theexampleinthechapterTheDefinitionFile,p.
24).
Insuchafilescalingfactorsettingsarecodedbytheinitialcharacters"SF"followedbythechannelnumber(twodigits)andanon-significantcharacter(e.
g.
doublepoint)andthescalingfactorvalue.
Thescalingfactorhastobeintherange1to9999.
Undersomecircumstancesyouneedtoadjustyourinputdatainordertogetagoodresolutionandscalingonthegraphicdisplay.
Ifyourdatarepresentsverycoarseorveryrefinedunitsyoucanmultiplyyouroriginaldatabyanyfactorbyusingthekeysor.
Avaluesmallerthanonewilldividetheoriginaldatawhichwillfitforacoarsedifferentiation.
Valueslargerthanoneincreasetheoriginaldatawhichshouldbedonewithveryrefinedinputdata.
Theadjustmentprocesswillnotaffectyourdata,anditoperatesonlyasanadaptationforthegraphicdisplay.
Again,youcanpresettheadjustmentfactorforeachchannelviaadefinitionfile(cf.
theexampleinthechapterTheDefinitionFile,p.
24)wheretheadjustmentfactorsettingsarecodedbytheinitialcharacters"AF"followedbythechannelnumber(twodigits)andanon-significantcharacter(e.
g.
adoublepoint)andtheadjustmentfactorvalue.
Thereisnolimitfortheadjustmentfactor(avalueequallingonehasnoeffect).
Withthehelpofthekey,,andyoumovethegraphinaverticaldirectionupanddown,respectively.
ScrollingtheCurrentDisplayTolookthroughthewholerecordingthekeysandmovethecurrentdisplayhalfascreenbackandforth,respectively,i.
e.
theformercentreofthescreenwillbeleftorrightinthenextscreen.
Bypressingthekeyyoumayjumptoanytimeposition(starttimeofthedisplay)withintherecordingrange.
PrintingaScreenTypingforcesECtodrawapixelimageofthescreen.
Thedrawingmaybedirectedtotheprinterortoaprintfile.
Theprintingprocessisdisplayedonthescreen(cf.
fig.
20)andmaybecancelledbypressing.
Youmaysetupyourprinterinthedefinitionfile(cf.
chapterThedefinitionfile,p.
24)oron-line(aftertypeandthentheprintercode;useforhelp).
Ifyouconfirmthepromptquestion(whichalsoshowstheprintermodel)youareaskedwhethertosendthescreenimagetoafileortotheprinter.
FilesarenamedSRCEEN00.
PRN,SCREEN01.
PRN,etc.
inregardofexistingprintfiles.
YoumayprintthefileslaterwiththescreenfileprintingutilityPSCREEN.
EXE.
Typing"PSCREEN[.
PRN][]ThedefaultisLPT1.
Example:"PSCREENscreen012"-printfileSCREEN01.
PRNatLPT2SelectionofResultsECwillindicatewithdifferentcolorswhetheraparticularresultisselectedfortheoutputresultfileornot.
Perdefaulttheselectedresultsarehighlighted,i.
e.
theyareinthesamecolorasthegraphline(youmaychangethecolorsettings,cf.
chapterChangingGraphicColorSettings,p.
33).
Dependingoncolumn10oftheentryscreentheresultsofachannelareeitherselectedofunselected(cf.
chapterTheUpperScreen,p.
24).
Youwillactivatearesultinthedisplaybypressing(leftarrowkey)or(rightarrowkey)whichshiftsbetweenadjacentresults.
Anactivatedresultishighlightedinwhiteinthelistontheleftaswellasinthegraph.
Pressingwillselectorde-selecttheactivatedresultfortheresultoutputfile(dependingontheformerstatusquo).
Anacousticsignalofeitherahigh(nowselec-ted)orlow(nowunselected)frequencyindicatesyouroperation.
Anactivatedresultcanbedeletedbypressing.
YouhavetoconfirmyourdecisionbecausetheactivatedresultwillberemovedfromthisECsessionanditcannotbereinserted-buttheoriginalinputresultfilewillnotbeaffected.
Notethatwhenworkingwithtimemarkingsresultsarenot(!
)implicitlyselectedwhenfittingthetimewindow.
Youhavetodothisexplicitlybyactivatingtherespectiveresultsandpressingwhichisfollowedbythehighfrequencytone.
Measuring:InsertingandRevisingResults1992Release20December2000E.
D.
A.
Manual1.
0432Fig.
21.
Measuringanevent:Zoomwindowwithcursor(righthand)andcorrespondingstatuspanel(lefthand).
Thekeysorareusedtoactivatethemeasuringpanelontheleftofthescreenaswellasasmallcrossinthemiddleofthescreen.
Thecrossbehaveslikeacursor:ifyoupresstheorkeysthecrossmovesandtheactualvalueandthecorrespondingtimearedisplayedinthemeasuringpanel(cf.
fig.
21).
Togetherwiththekeythecursormovementisspeededup.
Infact,whendetermininganamplitudeyouneedtospecifytwodatapoints,oneforthebasevalue(onset)andtheotherforthemaximumoftheampli-tude.
Pressingfixesonedatapointandmovestotheotherdatapointwhichisrepresentedbyasecondcross.
Again,youcanmovethesecondcrosswiththehelpofthecursorkeys.
Asbefore,willfixthisdatapointandjumpbacktothefirstdatapoint(andsoon).
Valuesandtimesofbothdatapointtogetherwiththeirdiffe-rencesareshownonthepanel.
Ifyouhavespecifiedanycompressionlargerthan1,ECwillzoomtheareaaroundtheactualdatapointwhenpressing.
Thisgivesyoutheopportunitytoselecteachdatavaluebymovinganewcrosswithinthezoomwindow.
Pressingorclosesthezoomwindowandshiftstotheotherdatapoint.
Afterchoosingtwodatapointsthekeyentersyourdecisionasachoiceforanewresult.
Ifyouconfirmyourdecisionafterpromptingtheresultisautomaticallyselectedfortheoutputresultfile.
Thekeywillcancelthemeasuring.
Sometimesyoudonotneedtwodatapoints,e.
g.
whenspecifyingtheR-wavepeakoranotherzeroamplitude.
Inthiscaseyousimplypressthekey(otherwiseyouhavetosetbothdatapointsexactlytothesameposition).
NotethatwhencorrectingECGresultsyoushouldnotmeasuretheR-waveamplitude.
CalculationsofRECAL.
EXErefertothedifferenceoftime(onsettimeofaparticularamplitude)minusrisetime(cf.
chapterCalculationsandmore,p.
36).
NormallyyoutaketheR-wavepeakasthereferencetime,andthereforetherisetimeshouldbezero.
Incontrasttothekeywhichwillalwaysinsertanewresultthekeywillmodifyanexistingresultifactivated.
InthiscaseyoumaychangethebasevalueandE.
D.
A.
Manual1.
041992Release03/04/199333themaximumoftheactivatedresult.
Thecursorwillstartatthebasevalueoftheparti-cularresultandswaptoitsmaximum.
TimeMarkingsIfyouareworkingwithatimefile(cf.
chapterTimeAnalysisandTimeFiles,p.
27)thecurrenttimemarker(numberandtime)isdisplayedatthetopleftcornerbetweenfilenameandchannellabel(cf.
fig.
17).
Allmanualresultselectionwillstorethistimeastheeventtime,i.
e.
theresultsofthe*.
Zfilesarerelatedtothismarkertime(cf.
example2inchapterListingaResultFile,p.
35).
Whenworkingwithtimemarkingsthekeywillactivatethecursorcrossinthemiddleofthetimewindow(whennoresultisformerlyactivated).
Youmayjumptodifferenttimemarkersbyeitherusingthekey(chooseanymarkernumberwithintherangeofdetectedmarkers)orthekeysandtoskiptothenextandprevioustimemarker,respectively.
Itisalsopossibletoinsertordeleteatimemarkermanually.
Thekeyinsertanewtimemarkeratthespecifiedtime.
Thekeydeletesthetimemarker.
Bothkeysadjustthetimefile.
Thekeystoaswellasthekeysareonlyfunctionalwhenusingatimefile.
ChangingGraphicColorSettingsThedefaultcolorsettingsofECmaybealteredbytheuser.
Pressingthekeytogglesbetweeneightdifferentbackgroundcolors(usingEGAorVGA).
Thekeyactivatesaleftpaneltomodifythecolorsofthecurrentlyactivatedchannel.
Thekeys,andtogglebetweenfifteenforegroundcolors(EGA/VGA)forthegraphline,theselected,and,theunselectedresults,respec-tively.
Thekeyquitsthechannelcolorpanel.
YouwillsavethecurrentcolorsettingswithtoafilenamedEC.
PAL.
WhenstartingtheprogramECsearchesinthecurrentDOSdirectoryforafileEC.
PALwiththecolorsettings.
1992Release20December2000E.
D.
A.
Manual1.
04344ResultsandCalculationsTheResultFileConvertingUtilityTheRECAL.
EXEroutineisamulti-purposeprogramtooperateonresultfiles.
Bytyping"RECAL][]switches:/afileconverting:ASCII=>binary(default:binary=>ASCII)/indexshowsamplerateindexofeventsinlisting(default:showeventnumber)/n[:]numberorevents**optionalfromtime/stattimerelatedstatistics(onlywith/n)**totime/p[:]IBI("hearttime")**inseconds/pk[:]IBIplausibilitycontrollowerHRlimit(interruptslisting)upperHRlimit(interruptslisting)/r[:]IBIphasicchanges(realtime)referencetime(default:timesfromtimefile*.
ZMD)calculationtime(default:1[sec])=>createsanewresultfile*.
R/rt:specificationsforIBIphasicchanges(realtime)calculationsbefore(default:5[sec])calculationsafter(default:10[sec])/zconverttimefile/thishelpRECALexpectsatleastaresultfilename()toconvertthebinaryinformationintoreadableASCIIcode.
AllASCIIoutputtothescreenmaybedirectedtoanASCIItextfileifyouspecifyanname.
ListingaResultFileExample1showsanEDAresultfilecreatedbyED.
ThecolumntypereferstothekindofSCR,i.
e.
1codesanormalSCR,2codesanSCRwithaninterpolatedrecoverytime/2(becauseofanewSCR),and0codesanSCRwithnorecoveryatall(recoverytime/2issetto-1.
000).
ThecolumntimeholdstherealtimeofthemaximumoftheSCRamplitude.
TheonsetoftheSCR(basevalue)isthistimeminustherisetime.
filename(binary)DEMO.
G01filename(ASCII)DEMO.
EDAthreshold:maximum0.
0250threshold:overlayedmaximum0.
0250threshold:speed(slope)0.
0750E.
D.
A.
Manual1.
041992Release03/04/199335percentageofmaximum=onset5.
00timeconstantinseconds0.
98samplerate200.
00Hzdelayinsamplepoints43.
00filterfrequency2.
00Hznumbertimetypeamplitudesloperiserecovery[sec]maximumtime/211.
30510.
4430.
6711.
2101.
08024.
36521.
3431.
7701.
3501.
38737.
08013.
7084.
9381.
2651.
11049.
84022.
2633.
6311.
1253.
393510.
99510.
3310.
7080.
7100.
325:2186.
35010.
1510.
2341.
0700.
7902291.
49011.
8782.
5001.
6951.
2352394.
13011.
3882.
3141.
0650.
8652497.
72512.
2913.
1071.
3551.
025okExample2showsanEDAoutputresultfilecreatedbyECwhileusingtimemarkings.
Onlyafewresultsfromtheinputresultfilehavebeenselected.
Thecolumnrecoverytime/2isreplacedbyeventtime,i.
e.
theactualreferencetimeofthetimefilewhenselectingaparticularresult.
Thecolumnslopemaximumisreplacedbylatency,i.
e.
timeminusrisetimeminuseventtime.
Astherehasbeenaninteractivecorrection(cf.
chapterMeasuring,p.
32)thereisa-1inthecolumntype.
filename(binary)DEMO.
Z01filename(ASCII)numbertimetypeamplitudelatencyriseevent[sec]time19.
84022.
2631.
5051.
1257.
210210.
99510.
3313.
0750.
7107.
210339.
660-11.
9133.
2251.
24535.
190462.
71010.
4852.
3101.
16059.
240Example3showsanECGresultfilecreatedbyED.
NotethattimeholdsthepeakoftheR-waveandthatthereisalwaysarisetimeofzero.
filename(binary)DEMO.
G04filename(ASCII)threshold:maximum0.
5000threshold:overlayedmaximum0.
5000threshold:speed(slope)0.
0750percentageofmaximum=onset5.
00timeconstantinseconds0.
00samplerate200.
00Hzdelayinsamplepoints0.
00filterfrequency0.
00Hznumbertimetypeamplitudesloperiserecovery[sec]maximumtime/210.
37591.
2160.
0000.
0000.
00021.
13091.
2300.
0000.
0000.
000:107.
20091.
2010.
0000.
0000.
000117.
91091.
2090.
0000.
0000.
0001992Release20December2000E.
D.
A.
Manual1.
0436OptionsYoumayusetheswitch/atoreconvertfromanASCIIfilegeneratedbyRECALtoabinaryfilebutthiswillonlyworkiftheASCIIlistingcontainstheheadentries,i.
e.
samplerate,etc.
Theswitch/indexwillshowthesamplerateindexofeacheventinsteadoftheeventnumber.
Example4showsatimefile(DEMO.
ZMD)listinggeneratedwiththeswitch/z.
17.
21000sec235.
19000sec359.
24000sec482.
21500secCalculationsTheswitches/n,/stat,/p,/rand/pkareusedtocalculateonaresultfile.
NumberofEventsTheswitch/ncountsthenumberofevents.
n=24resultsinDEMO.
G01.
.
.
Togetherwithatimerangespecificationonlytheeventsfittingthetimerangeareregarded.
Thecriticaltimeforfittingiscolumntimeminuscolumnrisetime.
Forexample,/n50.
0:100.
0calculatesona50secondsinterval.
n=73results(50.
000-100.
000sec=>87.
600n/min)inDEMO.
G04.
.
.
HeartPeriod,HeartRateandIBIThesamelogicholdsfortheswitch/pwhichcalculatesheartperiods(HP)andheartrate(HR)onIBIs,i.
e.
thetimebetweenadjacentresults.
IBItime[sec]HP[sec]HR[bpm]120.
8100.
7200083.
333221.
5600.
7500080.
000322.
2850.
7250082.
759423.
0150.
7300082.
192523.
7650.
7500080.
000624.
4800.
7150083.
916E.
D.
A.
Manual1.
041992Release03/04/199337725.
1450.
6650090.
226825.
7650.
6200096.
774926.
3650.
60000100.
0001026.
9700.
6050099.
1741127.
6200.
6500092.
3081228.
3600.
7400081.
0811329.
0650.
7050085.
1061429.
7600.
6950086.
33114IBI(n=15,20.
000-30.
000sec)inDEMO.
G04.
.
.
okPhasicHeartRateChangesTheswitch/rcalculatesonIBIsphasicchangesforrealtime.
Forexample,theswitch/r50.
0calculatesforthereferencetime50.
0secondsforstepsof1seconds(default)heartperiodandheartrate.
Additionally,ECcalculatesameanheartratefortheintervalbeforethereferencetimeandthedifferenceheartratetothemeanheartrate.
Interval[sec]HP[sec]HR[bpm][bpm]-545.
000-46.
0000.
7338500081.
761-446.
000-47.
0000.
7331500081.
839-347.
000-48.
0000.
6720000089.
286-248.
000-49.
0000.
6899000086.
969-149.
000-50.
0000.
6950500086.
325050.
000-51.
0000.
7305000082.
13685.
132151.
000-52.
0000.
6920750086.
6961.
564252.
000-53.
0000.
6580250091.
1826.
050353.
000-54.
0000.
7158500083.
816-1.
315454.
000-55.
0000.
7297500082.
220-2.
912555.
000-56.
0000.
7080500084.
740-0.
392656.
000-57.
0000.
7066500084.
908-0.
224757.
000-58.
0000.
6968000086.
1080.
976858.
000-59.
0000.
7041000085.
2150.
083959.
000-60.
0000.
7216000083.
149-1.
983Youmayuseatimefileasareferencetimeinput.
Ifthefirsttimeoftheswitch/requalszero(/ror/r0.
0:1.
0)RECALsearchesforatimefile*.
ZMD.
Ifthereisnofileyouarepromptedforatimefilename.
RECALwillcalculaterealtimephasicchangesforalltimesofthetimefile.
Togetherwiththeswitch/rtyoumaychangethedefaultsettingsofthepre-andpost-intervals.
Thenextexampleshowsalistingwiththeswitchsettings/r0:0.
5and/rt6:10whileDEMO.
G04servedastheinputresultfile.
timefile:DEMO.
ZMDInterval[sec]HP[sec]HR[bpm][bpm]17.
210-64.
210-4.
7100.
7750000077.
419-54.
710-5.
2100.
7564000079.
323-45.
210-5.
7100.
7495500080.
048-35.
710-6.
2100.
7800000076.
923-26.
210-6.
7100.
7771500077.
205-16.
710-7.
2100.
7737000077.
54907.
210-7.
7100.
7100000084.
50778.
0611992Release20December2000E.
D.
A.
Manual1.
043817.
710-8.
2100.
7430000080.
7542.
69328.
210-8.
7100.
7667500078.
2520.
19238.
710-9.
2100.
7900000075.
949-2.
11149.
210-9.
7100.
7826500076.
663-1.
39859.
710-10.
2100.
7750000077.
419-0.
641610.
210-10.
7100.
7515000079.
8401.
780710.
710-11.
2100.
7236000082.
9194.
858811.
210-11.
7100.
6924000086.
6558.
594911.
710-12.
2100.
7300000082.
1924.
131235.
190-632.
190-32.
6900.
7306000082.
124:Incaseoftheswitch/ranewresultfile*.
RisgeneratedbyRECALwhichmaybelaterinspectedwithRECAL.
ComparethefollowinglistingofthefileDEMO.
R04withthepreviousexample.
filename(binary)DEMO.
R04filename(ASCII)numbereventtimeNo.
off-timecal.
IBIHRmeanHR/[sec][sec][sec]deltaHR17.
210-64.
2100.
77577.
4190.
00027.
210-54.
7100.
75679.
3230.
00037.
210-45.
2100.
75080.
0480.
00047.
210-35.
7100.
78076.
9230.
00057.
210-26.
2100.
77777.
2050.
00067.
210-16.
7100.
77477.
5490.
00077.
21007.
2100.
71084.
50778.
06187.
21017.
7100.
74380.
7542.
69397.
21028.
2100.
76778.
2520.
192107.
21038.
7100.
79075.
949-2.
111117.
21049.
2100.
78376.
663-1.
398127.
21059.
7100.
77577.
419-0.
641137.
210610.
2100.
75279.
8401.
780147.
210710.
7100.
72482.
9194.
858157.
210811.
2100.
69286.
6558.
594167.
210911.
7100.
73082.
1924.
1311735.
190-632.
1900.
73182.
1240.
0001835.
190-532.
6900.
73581.
6330.
0001935.
190-433.
1900.
74280.
9170.
000:TimeRelatedIBIStatisticsTheswitch/statcalculatesseveraltimespecificstatisticsforIBIs,i.
e.
mean(MEAN),standarddeviation(STDEV),variancecoefficient(VARCOEF),meansumofsquaresofsuccessivedifferences(MSSD),instantaneousheartrate(IHR).
ForadetailedexplanationofthesestatisticalparametersandtheirrelevancewerecommendthearticleofvanDellen,Aasman,Mulder&Mulder(1985)asareference.
Incontrasttotheswitch/nallcalculationsrefertothetimeintervalbetweenthefirstandthelastheartbeat.
Forclarificationlookatthefollowingthreeexamples(correspondingcommandsareenteredaftertheDOSprompt):E.
D.
A.
Manual1.
041992Release03/04/199339C:\EDA\DEMO>recaldemo.
g04/n15:75/statn=86events(15.
000-75.
000sec=>86.
000n/min)inDEMO.
G04.
.
.
n=85IBI(15.
435-74.
660sec=>86.
112n/min)MEAN=696.
7647058824msecSTDEV=43.
9338374472msecVARCOEF=6.
3054051212MSSD=1245.
8333334122IHR=1.
4352047277HzokC:\EDA\DEMO>recaldemo.
g04/n15:75n=86events(15.
000-75.
000sec=>86.
000n/min)inDEMO.
G04.
.
.
okC:\EDA\DEMO>recaldemo.
g04/n/statn=143eventsinDEMO.
G04.
.
.
n=142IBI(0.
375-99.
925sec=>85.
585n/min)MEAN=701.
0563380281msecSTDEV=45.
4794435129msecVARCOEF=6.
4872737105MSSD=1128.
5460992791IHR=1.
4264188850HzokPlausibilityCheckTheswitch/pkisusedtochecktheIBIsforextrabeats(systoles)orartifacts.
Forexample,theswitch/pk55:85willlistallIBIsandtheirrespectiveHRbutwillpausethelistingwheneveranIBIofanHRbelow55bpmorover85bpmappears.
C:\EDA\DEMO>recaldemo.
g04/pk55:85IBIHP[sec]HR[bpm]11.
1300.
7550079.
47021.
8550.
7250082.
75932.
5750.
7200083.
33343.
3250.
7500080.
00054.
1250.
8000075.
00064.
9000.
7750077.
41975.
6450.
7450080.
53786.
4250.
7800076.
92397.
2000.
7750077.
419107.
9100.
7100084.
507118.
6750.
7650078.
431129.
4650.
7900075.
9491310.
2400.
7750077.
4191410.
9900.
7500080.
0001511.
6800.
6900086.
9571992Release20December2000E.
D.
A.
Manual1.
0440--------1612.
4100.
7300082.
1921713.
1800.
7700077.
9221813.
9500.
7700077.
9221914.
6950.
7450080.
5372015.
4350.
7400081.
0812116.
1250.
6900086.
957--------:ResultFileStructureExperiencedprogrammerscanusetheresultfilescreatedbyED.
EXE,EC.
EXEandRECAL.
EXEtocalculatetheirownparameters.
ForthisreasonthestorageformatoftheeventsintheresultfileisgivenasapartofaBORLANDTurboPascalsourcecodeinthefileRESULT.
PAS.
Regardthecommentsinthesourcecode!
{RESULT.
PAS(C)1992H.
-J.
Trosiener&J.
Kayser}{ThefollowingdeclarationsrefertothefilestructureofresultfilescreatedbyED.
EXE,EC.
EXEandRECAL.
EXE.
Thefirstrecordofaresultfileconsistsof=2,thesecondrecordof=3,andallfollowingrecordsareof=1.
Dependingonthekindofthevariablesof=1differfromthemeaningofthevariablenamewhichholdonlyfor=NormalType.
}TYPECodeType=(NormalType,SlopeType,LatencyType,IBI_RealtimeType);ResultType=RECORDCASEINTEGEROF1:(Index:LongInt;TimePoint:REAL;AmplitudeType:INTEGER;Amplitude:REAL;SlopeMaximum:REAL;RiseTime:REAL;RecoveryTime:REAL);2:(Min_Ampl_Crit:REAL;Min_Inter_Top:REAL;Min_Slope_Crit:REAL;Percentage:REAL;Code:CodeType);3:(TimeConstant:REAL;SampleRate:REAL;FilterDelay:REAL;Frequency_low:REAL;Frequency_high:REAL);END;E.
D.
A.
Manual1.
041992Release03/04/19938TheCODASpackage(Computer-basedOscillographandDataAcquisitionSystem)isasoftwareandhardwareutilitythatturnsapersonalcomputerintoahighperformancedataacquisitionsys-tem(DataqInstruments,Inc.
,1988).
41VARResultFile:FILEofResultType;5ModifyingCODASDataFilesTheCOSAM.
EXEroutineallowsyoutoreducetheoriginalsamplerateofaCODAS8datafileinordertoreducetheamountofdata,toextractpartsfromtheoriginalfile,and,tojoindifferentpartstoanewCODASdatafile.
Bytyping"COSAM/"youwillgetacommandhelpscreen.
CODASFileOperatingUtilityVersion1.
0401/07/1993Syntax:COSAM[][][]arguments:/rinteractivedatareductiontoasmallersamplerate/eextractparts/jjoinwithsecondCODASinputfile/iinformationaboutCODASinputfile/thishelpCODASinputfileCODASoutputfilesecondCODASinputfileforjoiningIfyouarecreatinganewCODASdatafile,i.
e.
usingoneofthearguments/r,/e,or/j,thenthecalibrationfile*.
CALoftheCODASinputfileiscopied.
1992Release20December2000E.
D.
A.
Manual1.
0442ReferencesBoucsein,W.
(1988).
ElektrodermaleAktivitt:Grundlagen,MethodenundAnwendun-gen.
Berlin:Springer.
(Englishtranslation1992:ElectrodermalActivity.
NewYork:PlenumPress.
)CookIII,E.
W.
&Miller,G.
A.
(1992).
DigitalFiltering:BackgroundandTutorialforPsychophysiologists.
Psychophysiology,29,350-367.
DataqInstruments,Inc.
(1988).
CODASUser'sManual.
SoftwareReleaseLevel4.
ManualRevisionH.
825SweitzerAve.
,Akron,Ohio44311,USA.
vanDellen,H.
J.
,Aasman,J.
,Mulder,L.
J.
M.
&Mulder,G.
(1985).
TimeDomainversusFrequencyDomainMeasuresofHeart-RateVariability.
InOrlebeke,J.
F.
,Mulder,G.
&Doornen,L.
J.
P.
van(Eds.
),PsychophysiologyofCardiovascularControl,pp.
353-374.
NewYork:PlenumPress.
Elbert,T.
&Rockstroh,B.
(1980).
SomeRemarksontheDevelopmentofaStandar-dizedTimeConstant.
Psychophysiology,17,504-505.
Martin,I.
&Venables,P.
H.
(1980).
TechniquesinPsychophysiology.
Wiley:Chiches-ter.
Thom,E.
(1988).
DieHamburgerEDA-Auswertung.
InBoucsein,W.
(Hrsg.
),Elektro-dermaleAktivitt:Grundlagen,MethodenundAnwendungen,S.
501-514.
Berlin:Springer.
Tietze,U.
&Schenk,C.
(1978).
Halbleiter-Schaltungstechnik(4.
Aufl.
).
Berlin:Springer.
Trosiener,H.
-J.
&Kayser,J.
(1992).
AnalysisandInteractiveControlofElectrodermalandCardiacActivity:AComputerProgramforIBMPC'sandCompatibles.
Posterpresentedatthe20thannualmeetingoftheDGPA,June18-20,Ham-burg.
E.
D.
A.
Manual1.
041992Release03/04/199343AppendixA.
TheDefinitionFileED.
DEFcod;typeofData(-File):COD,BIN,ASCecg1;typeofAnalysis(eda,ecg1,ecg2,pow,int)-5.
0;lowerborderofsignalrange5.
0;upperborderofsignalrange0;timeconstantinseconds,0=notimeconstantbessel;filtertype(forfutureversions)notch;filtercharacteristic(Low,High,Band,Notch(Bandrejection),none)2;orderofthefilter(2,4,6,8,10)48;uppercutoff-FrequencyforLP/HPfilter;lowerforBandfilters/notch52;uppercutoff-frequencyforbandfiltersandbandrejectionnone;binaryOutput(tcor,filo,filc,orig,none).
4;Threshold:maximum(smallestamplitudeofamaximum)0;Threshold:overlayedmaximum0;Threshold:speed0;Threshold:acceleration0;frequencyofr-waveforenhancement/ortimeconstantofintegrator(OnlyformethodsECG2andINT)0;segmentlength(DataPoints)inEMS/RAM/disk(0=automaticuse)B.
TheDefinitionFileEC.
DEF;EC.
DEF-parametersettings(definitions)ofEC.
EXE(C)Kayser01/07/1993;EC.
EXEsearchesforafileEC.
DEF(anotherfilenamemaybe;specifiedasthesecondcallparameterwhenstartingtheprogram).
;Thisfileisanexampleofaparametersettingfile:;-emptylinesareignored;-commentsstartwithasemicolon;-theorderofallparametersfortheheadsettingsisfixed;-headsettingsoccuronlyonceandrefertoallchannels(thedefault;iscodedbyaminussign);-sequencedsettingsmaybeleft,i.
e.
itispossibletohavenohead;settingsatall(defaultsettings);-headsettings7;rateofcompression(defaultis1,notsmallerthan1,;wholenumber)y;eventrelatedanalysis(usingtimemarkings)-;filenameoftimemarkings(-=*.
ZMD)5;channelnumberfortimeanalysis1.
0;startoftimewindow[sec]inrelationtoatimemarker5.
0;endoftimewindow[sec]inrelationtoatimemarker1.
0;heightofmarkings(abaseofzeroissupposed)1992Release20December2000E.
D.
A.
Manual1.
04440;duration[sec]ofthe1strectangularimpulse(0=anytime)0.
5;duration[sec]betweenbothimpulses(0=anytime)1.
0;duration[sec]ofthe2ndrectangularimpulse(0=anytime)2;timemarkinginpattern(0=riseedgefirstrectangular;impulse,1=decayedgefirstrectangularimpulse,2=rise;edgesecondrectangularimpulse,3=decayedgesecond;rectangularimpulse;ifeachdurationequalszeroaneven;number(0or2)willleadtoadetectionofallriseedges,;anoddnumber(1or3)willdetectalldecayedges);-adjustmentfactorsettings;AF02:0.
01;adjustmentfactorchannel02(datadividedby100);AF03:0.
1;adjustmentfactorchannel03;-scalingfactorsettings(default=autoscaling)arecodedbythe;initialcharacters"SF"followedbythechannelnumber(twodigits);andanotsignificantcharacter(e.
g.
doublepoint),i.
e.
the;scalingfactorvaluemustbeoncolumn6-15;-scalingfactorsettings(range1to9999);SF01:100;scalingfactorchannel01;SF02:100;scalingfactorchannel02;SF03:200;scalingfactorchannel03;SF04:50;scalingfactorchannel04;-parametersettingsofaparticularchannel(channelentries)are;introducedby'CHANNEL[n]'([n]=channelnumberwithtwodigits,;e.
g.
'01',max.
8channelsarepossible);-entriesforaparticularchannelareoptional;ifachannelis;mentionedallentriesforthatchannelhavetobeexplicit;-thedefaultiscodedbyaminussignCHANNEL01;startentriesofchannel1y;y=showdataofchannel,n=ignoredataofchannelSCRleft;label(commentmax.
12characters)-;binaryfilename(-=*.
D01)y;y=usebinarydata,n=donotusebinarydata-;resultfilename(-=*.
G01)y;y=showresults,n=donotshowresults-;filenameforcorrectedresults(-=*.
K01)y;y=correctresults,n=donotcorrectresultsn;selectionofresults(y=all,n=none)CHANNEL02;startentriesofchannel2y;y=showdataofchannel,n=ignoredataofchannelSCRright;label(commentmax.
12characters)-;binaryfilename(-=*.
D01)y;y=usebinarydata,n=donotusebinarydata-;resultfilename(-=*.
G01)y;y=showresults,n=donotshowresults-;filenameforcorrectedresults(-=*.
K01)y;y=correctresults,n=donotcorrectresultsn;selectionofresults(y=all,n=none)CHANNEL03;startentriesofchannel3E.
D.
A.
Manual1.
041992Release03/04/199345y;y=showdataofchannel,n=ignoredataofchannelEOG;label(commentmax.
12characters)-;binaryfilename(-=*.
D01)n;y=usebinarydata,n=donotusebinarydata-;resultfilename(-=*.
G01)n;y=showresults,n=donotshowresults-;filenameforcorrectedresults(-=*.
K01)y;y=correctresults,n=donotcorrectresults-;selectionofresults(y=all,n=none)CHANNEL04;startentriesofchannel4y;y=showdataofchannel,n=ignoredataofchannelECG;label(commentmax.
12characters)-;binaryfilename(-=*.
D01)y;y=usebinarydata,n=donotusebinarydata-;resultfilename(-=*.
G01)n;y=showresults,n=donotshowresults-;filenameforcorrectedresults(-=*.
K01)y;y=correctresults,n=donotcorrectresultsy;selectionofresults(y=all,n=none)CHANNEL05;startentriesofchannel5y;y=showdataofchannel,n=ignoredataofchannelmarkings;label(commentmax.
12characters)-;binaryfilename(-=*.
D01)n;y=usebinarydata,n=donotusebinarydata-;resultfilename(-=*.
G01)n;y=showresults,n=donotshowresults-;filenameforcorrectedresults(-=*.
K01)n;y=correctresults,n=donotcorrectresults-;selectionofresults(y=all,n=none);theentry'PRINTER[n]'initializestheprinter([n]=printernumber;withthreedigits,e.
g.
'001');thedefaultsettingisastandardprinter;uptonowthefollowingprinteraresupported:;001-EpsonMX100;002-EpsonRX-80,Epson80F/T;003-EpsonLQ-850/1050;004-IBM4207-002andcompatibleIBMprinters;005-Hewlett_PackardLaserJetII,Hewlett_PackardDeskJetPLUS;006-PanasonicKX-P1092;-printerPRINTER004;initializesanIBMprinterC.
AnEDAResultFileListingExamplefilename(binary)DEMO.
G01filename(ASCII)DEMO.
EDAthreshold:maximum0.
0250threshold:overlayedmaximum0.
0250threshold:speed(slope)0.
0000percentageofmaximum=onset5.
00timeconstantinseconds1.
00samplerate200.
00Hz1992Release20December2000E.
D.
A.
Manual1.
0446delayinsamplepoints43.
00filterfrequency2.
00Hznumbertimetypeamplitudesloperiserecovery[sec]maximumtime/211.
29510.
4410.
6681.
2351.
08024.
36021.
3311.
7621.
3501.
37337.
07013.
6754.
9101.
2551.
10549.
83522.
2543.
6301.
1203.
249510.
99010.
3240.
6980.
7100.
325616.
79026.
3647.
3261.
61020.
091725.
84524.
5845.
8431.
4902.
174827.
46511.
3112.
7760.
7750.
500936.
64510.
0540.
0801.
0500.
5801039.
67512.
0432.
7532.
1001.
2651142.
36511.
5762.
8120.
9650.
7301244.
47511.
6702.
9390.
9750.
7001349.
16011.
5332.
8170.
9350.
6901451.
27510.
4160.
9220.
7400.
4301553.
64011.
8212.
7251.
1600.
7851662.
70510.
4820.
6621.
1701.
4851767.
98001.
4412.
1291.
195-1.
0001868.
97011.
6352.
6490.
9850.
6751975.
51012.
4982.
8221.
5201.
0052077.
68010.
9812.
0150.
8250.
6502181.
89510.
2790.
5190.
9450.
6702286.
35010.
1540.
2341.
1350.
8002391.
48011.
8602.
4851.
6901.
2202494.
12511.
3812.
3071.
0600.
8652597.
72012.
2743.
0961.
3551.
020E.
D.
A.
Manual1.
041992Release03/04/199347Index3030323032,335,2624,2626303030303030,3126,3031,3231,3226,3030,3130,3323,26,28,30-336,23,25,29,3128,30293030303033333330,3332,333333333332,3333acceleration8,11,13,43actualtime29adjustmentfactor30,44amplitudecriterion1,15,17arbitraryscalingfactor29ASCII1,2,4-7,10,13,23,24,28,34-36,38,45BAND1,9,10,13,21,43Band-pass9,10basevalue25,32-34batchmode6,10,13-15Besselfilter9,11,15binarydata5,10,23-25,44,45binaryoutput1,8,10,11,13,43BORLAND40bpm39calculations2,13,28,33,34,36,38calibrationfile3,4,41channelnumber5,8,14,24,30,43,44co-processor2,3CODAS1-4,6,8,23-26,41,42compression8,12,15,19,22,26,27,32,43correctionmode27COSAM.
EXE2,3,41cut-offfrequency10dataformat1,4datapoint4,26,32DataqInstruments,Inc.
41,42datatype6,8definitionfile1,13,14,23,24,26,27,30,31,43demofiles3,24displacement8,11dottedlines27duration27,43,44E.
D.
A.
i,1,3,5EC2-5,13,23-35,37,40,43EC.
DEF3,23,24,43EC.
EXE2-5,23,24,40,43ECG1,3,6,7,18-21,25,33,35,45ECG17,13,14,18-20,43ECG27,12,13,18,20,43ED1-6,8,11,13,14,23,25,34,35,40,43ED.
DEF1,3,13,14,43ED.
EXE1-5,14,23,40EDA1,3,6-8,11,13,15,16,15-17,34,35,39,42,43,45edge27,44electrocardiogram18electrodermalactivity1,7,15,42EMG1,3,9,21,22,21,22EMS2,14,15,43entryscreen3,5,6,8,23-28,31eventdetectioni,1,5,6,8,13,21events1,2,5,6,11,13,23,25-27,34,36,39,40filename4-6,8,14,23-27,29,33-35,37,38,43-45filetype4filter1,8-13,15,20,21,35,43,45filtercharacteristic9,13,43filterorder10filtertype9,13,43frequency4,10-13,17,18,24,25,31,32,35,40,42,43,45graphic2,3,15,16,19,22-31,33graphiccolors16,19,22groupdelay9-11HAMM12Hamming12HANN12Hanning12hardwarerequirements2heartperiod1,36,37heartrate1,27,36-38hearttime34help3,6,8,23,25-27,29-32,34,41HIGH1,9-13,31,32,40,41,43horizontalposition30IBI34,36-40IHR38,39INT7,13,43integration1,3,7,21,221992Release20December2000E.
D.
A.
Manual1.
0448interactivemode6,10,12,14,15,18,23label24,25,29,33,44,45LOW1,8-10,13,17,31,40,43manualscaling12maximum7,8,11-13,15-20,25,26,29,32-35,43,45meansumofsquaresofsuccessivedifferences38memory2,14,24method6-8,11,13,14,16-19,21,22MSSD38,39NOTCH1,9,10,13,21,43on-linehelp3,6,23,26,29onset1,32-35,45openfiles24,26parameterpre-definitions13Parzen12phasicchanges34,37POW7,13,43powerspectrum1,3,7,11,20,21precision28printfile30,31printer24,30,31,45PSCREEN.
EXE2,3,31R-wave7,12,13,18,19,32,33,35,43R-waveenhancement12,18RAM2,3,14,43range4,7,8,12,13,21,26,29,30,33,36,43,44rangeoverflows13RECAL.
EXE2,3,28,33,34,40recoverytime16,17,20,34,35RECT12rectangularwindow12,21resultfile1,3,5,13,25,30-38,40,44,45RESULT.
PAS3,40risetime17,33-36samplerate2,4,8,18,21,24,25,28,34-36,41,45scaling12,21,29,30,44scalingfactor29,30,44SCR1,17,18,25,34,44scrolling30segment12,14-16,19,21,43selection11,25,31,33,44,45speed2,8,11,13,35,43,45STDEV38,39systemfiles3tc9,21,22threshold8,11,13,34,35,43,45timeanalysis26-28,33,43timeaxis29timeconstant1-3,5,7-10,13,21-23,35,42,43,45timefile3,5,26-28,33-37timemarkings1,2,5,7,26,28,31,33,35,43timewindow2,27,31,33,43totaltime25TurboPascal3,40VARCOEF38,39variancecoefficient38virtualmemory14zeroamplitude32zoomwindow32