Lub hlwb kev sib txuas hloov pauv thaum pw tsaug zog los ntawm kaw-voj Transcranial Neurostimulation Predict Metamemory rhiab heev Part 2

Electrophysiological Recordings thiab Transcranial Stimulation

Cov ntaub ntawv EEG tau sau thiab kev txhawb nqa tau siv ib txhij nrog tus qauv StarStim64 los ntawm Neuroelectrics, Inc., nrog rau 64-channel neoprene lub taub hau hau nrog rau 32 EEG channels thiab 32 stimulation channels.

Kev txhim kho txuas ntxiv ntawm kev tshawb fawb thiab thev naus laus zis niaj hnub no tau ua rau tib neeg txoj kev tshawb fawb ntawm lub hlwb ntau thiab ntau qhov tob. Nyob rau hauv xyoo tas los no, ntau cov kev tshawb fawb tau pom tias muaj kev sib raug zoo ntawm electroencephalography (EEG) thiab kev nco. EEG yog ib txoj hauv kev kaw lub hlwb hluav taws xob ua haujlwm. Nws qhia txog tib neeg lub hlwb hluav taws xob teeb liab nyob rau hauv daim ntawv ntawm kab kos los ntawm kev kuaj xyuas lub hlwb cov teeb liab bioelectrical. Kev nco yog lub peev xwm ntawm lub hlwb neuron system kom tau, txheej txheem, khaws cia, paub txog, thiab rov qab cov ntaub ntawv. Kev sib raug zoo ntawm ob leeg tuaj yeem hais tau zoo sib xws.

EEG sau lub cev txav mus los, lus, kev xav, thiab cov ntaub ntawv xav paub, thiab tau siv dav hauv kev lag luam xws li tshuaj, neuroscience, thiab psychology. Ntawm lawv, cov kws tshawb fawb pom tias EEG muaj kev sib raug zoo nrog kev nco. Cov ntaub ntawv EEG tuaj yeem pab peb nkag siab zoo txog cov txheej txheem ntawm kev tsim thiab khaws cia, thiab muab cov ntaub ntawv tseem ceeb rau kev kho mob ntawm kev nco. Raws li kev tshawb fawb, kev tsim lub cim xeeb ntev yog nrog los ntawm kev siv hluav taws xob sib txawv thiab kev faib tawm, thiab thaum tib neeg rov nco qab, EEG cov cim tseem yuav hloov pauv, ntxiv kev lees paub qhov ze ntawm EEG thiab kev nco.

Tsis tas li ntawd, EEG kuj tseem tuaj yeem pab cov kws tshawb fawb nkag siab txog ntau hom kev nco, xws li kev nco luv luv, kev nco mus ntev, kev nco qab, thiab kev nco txog kev xav. Hauv kev tshawb fawb, EEG tsom xam tuaj yeem taw qhia cov haujlwm hluav taws xob los ntawm ntau lub hlwb. Cov haujlwm tshwj xeeb hluav taws xob no tuaj yeem qhia tau tias cov ntaub ntawv nco tau raug xa mus thiab ua tiav li cas hauv lub hlwb thiab tseem qhia txog qhov ua rau kev nco tsis zoo.

Nws tuaj yeem hais tias EEG muaj qhov tseem ceeb rau kev txhim kho lub neej zoo. Cov neeg niaj hnub no siv cov khoom siv hluav taws xob xws li khoos phis tawj thiab xov tooj ntawm tes tau ntev thiab tseem vam khom ntau dhau ntawm kev tshaj xov xwm, uas ua rau tib neeg nyuaj rau mloog thiab tsis nco qab tau yooj yim. Kev tshawb fawb qhia tau hais tias cov lus qhia luv luv EEG tuaj yeem txhim kho tib neeg txoj kev kawm muaj peev xwm thiab kev nco. Los ntawm kev kho tus kheej EEG, qhov txawv txav hauv lub hlwb tuaj yeem kho tau, lub hlwb ua haujlwm tau zoo dua qub, thiab kev nco thiab kev kawm tau zoo dua qub.

Hauv cov ntsiab lus, EEG thev naus laus zis yuav tsum tau siv dav hauv peb lub neej niaj hnub, thiab tseem xav tau kev tshawb fawb ntxiv los tshawb nrhiav ntau ntawm nws cov ntawv thov thiab muaj peev xwm. Nyob rau yav tom ntej, kev txhim kho txuas ntxiv ntawm EEG thev naus laus zis yuav ua rau peb muaj cib fim thiab kev sib tw ntau dua, ua rau peb muaj lub hlwb zoo dua thiab muaj kev noj qab haus huv. Nws tuaj yeem pom tias peb yuav tsum txhim kho kev nco, thiab Cistanche deserticola tuaj yeem txhim kho kev nco, vim Cistanche deserticola muaj antioxidant, tiv thaiv kev mob, thiab tiv thaiv kev laus, uas tuaj yeem pab txo qis oxidation thiab inflammatory tshwm sim hauv lub hlwb, yog li tiv thaiv cov kab mob. kev noj qab haus huv ntawm lub paj hlwb. Tsis tas li ntawd, Cistanche deserticola kuj tseem tuaj yeem txhawb kev loj hlob thiab kho cov paj hlwb, yog li txhim kho kev sib txuas thiab kev ua haujlwm ntawm neural networks. Cov teebmeem no tuaj yeem pab txhim kho kev nco, kev kawm, thiab kev xav nrawm, thiab tseem tuaj yeem tiv thaiv kev loj hlob ntawm kev paub tsis meej thiab kab mob neurodegenerative.

ClickKnoww txhim kho lub sijhawm luv luv

EEG channels tau muab tso rau hauv ib qho montage raws li 10-10system. EEG cov ntaub ntawv tau sau los ntawm 23 ntawm 32 qhov chaw ua piv txwv ntawm 500 Hz thiab hais txog tochannel Fz. Qhov ntxiv 9 electrodes tau kaw electrooculogram, electromyogram, thiab electrocardiogram kom tso cai rau kev tshawb pom cov theem pw tsaug zog. Tsis muaj kev lim dej hauv online, tshwj tsis yog linenoise (60 Hz) filtering, tau siv thaum sau.

Active transcranial stimulation tau xa nyob rau hauv daim ntawv ntawm cov cim arrays ntawm spatiotemporalcurrents thoob plaws 32 stimulation electrodes. Tag nrho cov hluav taws xob zoo tam sim no tau teem rau qis dua 2.5 mA, nrog rau txhua qhov channel tam sim no hauv qab 1.5 mA thiab tshaj 150 uA.

Cov qauv ntawm kev txhawb nqa tau tsim los ntawm gradient qhovntsej thiaj tsis zoo kom ua kom zoo tshaj plaws orthogonality thoob plaws lub zog hluav taws xob hauv 3-D ntim ntawm tus neeg laus lub taub hau template bypenalizing correlations thiab anticorrelations.

Thaum lub sij hawm kev sim, rau lub Active mob, txhua tus ntawm 14 vignettes yog random txuam los yog "tagged" nrog ib tug txawv qauv ntawm stimulation, thiab kev cai templates rau txhua tus qauv yog programmed nyob rau hauv lub Neuroelectricsstimulation tswj software, CoreGUI. Tsuas yog ib nrab ntawm cov qauv tES no tau rov qab siv dua thaum hmo ntuj rau cue nco reactivation (Tag & Cue), thiab kev soj ntsuam tus cwj pwm tsom rau kev nco qab cov lus teb rau cov phiaj xwm vignettes.

Cov qauv tshwj xeeb tau tsim los ua kom nco txog qhov tshwj xeeb, tsis yog hloov pauv kev ua haujlwm ntawm neural hauv kev tshwj xeeb; Yog li, lub hlwb cov lus teb thiab kev sib txuas hloov pauv mus rau ib tus neeg STAMPs tsis tau txheeb xyuas thiab hloov pauv tau ua ke kom nce mus sib hais.

Slow-Wave Oscillation Detection

Stimulation tau cia li xa lub sij hawm-xauv mus rau cov xeev ntawm SWOs nrog kaw-loopdetection algorithm, yoog los ntawm ib tug yav tas los tsim algorithm (Cox li al., 2014). Thaum pw tsaug zog, cov ntaub ntawv EEG tuaj los ntawm 13 pem hauv ntej-parietal-central channels (Cz, FC1, FC2, CP1, CP2, Fz, C4, Pz, C3, F3, F4, P3, thiab P4) tau khaws cia rau hauv kev khiav haujlwm {{17} }s tsis tu ncua hloov tshiab.

Txhawm rau ntxuav cov ntaub ntawv, txav qhov nruab nrab rho tawm nrog lub qhov rais 1-s tau siv, thiab txhua txoj kev tshaj li 500 μV min-rau-max amplitude thoob plaws lub sijhawm tsis raug tso tseg.

Cov channel tseem tshuav tau nruab nrab ua ke los tsim cov channel virtual rau SWOdetection muaj zog. Lub zog nyob rau hauv lub suab qeeb qeeb (0.5–1.2 Hz) tau muab piv rau tag nrho broadbandpower (0.1–250 Hz), thiab yog tias qhov piv txwv no ntau dua 0.3, ces lub ncov meej ntawm lub hwj chim nyob rau hauv lub suab qis-yoj band (qhov nruab nrab zaus) raug xam.

Cov channel virtual tau lim nrog qhov kev txiav txim thib ob Butterworth bandpass lim nrog 1-Hz bandwidth nyob ib ncig ntawm qhov nruab nrab zaus (nrog qhov tsawg kawg nkaus txiav tawm ntawm 0.1 Hz), thiab Hilbert hloov tau siv. Thephase tau hloov rov qab 90 degrees, thiab cov kev xav txog kev xav tau muab rho tawm, yielding theinstantaneous theem.

Lub sine yoj ntawm qhov nruab nrab zaus nrog rau theem optimized, offset, thiab amplitude ntawm cov teeb liab lim tau raug tsim tawm. Lub sine yoj tau npaj rau pem hauv ntej rau lub sijhawm los txiav txim siab lub ntsiab lus tom ntej ntawm xoom theem (piv txwv li, qhov pib ntawm lub xeev), thiab kev txhawb nqa tau pib thaum lub sijhawm no taw tes thiab kav ntev rau qhov kev kwv yees nce-xeev nrog ramp-up thiab ramp-down. ua 100ms.

Thaum lub sij hawm Sham hmo ntuj, lub sij hawm stimulation ntsiab lus raug kaw, tab sis tsis muaj stimulation raug xa. Yog li, qhov kev sib piv ntawm Active thiab Sham hmo yog nyob rau tib lub sij hawm cov ntsiab lus ntawm SWO, nrog rau qhov sib txawv tsuas yog kev xa tawm ntawm STAMP stimulation.Validation ntawm up-state detection tau ua rau cov ntaub ntawv los ntawm Sham hmo ntuj kom tsis txhob artifactsproduced los ntawm STAMPsonn Active hmo.

Cov cim tau raug muab rho tawm los ntawm lub sijhawm cov ntsiab lus ntawm qhov kev kwv yees los ntawm sine yoj haum. Qee qhov xwm txheej, vim tias kho vajtse qeeb, lub sijhawm stimulation tau txav mus rau hauv lub xeev tom qab kuaj pom tam sim ntawd, uas tuaj yeem ua rau muaj kev hloov pauv ntawm theem mus rau qee qib. Epochs lub sij hawm kaw rau stimulation pib cim (−5 rau +5 s) tau muab rho tawm los ntawm raw EEG cov ntaub ntawv thiab bandpass lim nyob rau hauv lub 0.5–1.2 Hz ntau.

Phasevalues ​​ntawm txhua qhov pib cim lub sij hawm taw tes tau suav nrog kev hloov pauv Hilbert. Qhov ntsuas theem nruab nrab ntawm kev sim tau suav rau txhua tus neeg koom nrog, thiab cov theem nruab nrab no tau xa mus rau qhov kev xeem v-test rau qhov sib txawv ntawm 0 degrees (Berens, 2009). Cov duab thiab cov txheeb cais ntawm kev siv tau ntawm qhov kaw-voj algorithm tau nthuav tawm hauv Cov Ntaub Ntawv Txhawb Nqa.

EEG Data Processing thiab Graph Theoretic Analysis

EEG cov cim tau raug txheeb xyuas off-line nrog EEGLAB thiab kev cai MATLAB scripts. Txhawm rau ntxuav cov ntaub ntawv thiab cais cov suab nrov ntawm lub paj hlwb teeb liab, EEG cov ntaub ntawv tau qis-sampledto 250 Hz, xa mus rau qhov nruab nrab ntawm tag nrho cov electrodes, thiab bandpass lim siv aButterworth lim (0.1–70 Hz, 24 dB / oct).

Ocular thiab nonneural artifacts raug txheeb xyuas thiab tshem tawm manually siv txoj kev tshawb fawb ywj pheej (ICA) hauv EEGLABand nws cov infomax algorithm (Delorme & Makeig, 2004).

Cov ntaub ntawv tsis pub dawb tau muab faib ua ntu hauv ntu uas muaj −6.4 txog −2.4 ua ntej qhov pib ntawm stimulation raws li lub hauv paus qhov rais, thiab 3 mus rau 7 s tom qab offset ntawm stimulation raws li lub qhov rais tsom xam. Rau txhua lub sijhawm, lub zog spectrum ntawm txhua channel tau suav los ntawm kev siv Welch spectralestimation txoj kev ntawm 2048-ms ntu nrog 1,024-ms (50%) sib tshooj, nrog rau txhua ntu los ntawm lub qhov rais Hamming.

Kev sib txuas ua haujlwm thiab cov duab theoretical tsom xam tau ua nrog BioNeCtToolbox, kev cai MATLAB toolbox rau kev txheeb xyuas lub hlwb (Casanova, El-Baz, & Suri, 2017). Kev ua haujlwm ntawm lub hlwb kev sib txuas ntawm cov hlwb hauv cheeb tsam yog kwv yees los ntawm kev suav cov kev sib raug zoo hauv qhov zaus ntawm EEG cov cim (Bowyer, 2016; Mohammad-Rezazadeh li al., 2016; Nolte et al., 2004).

Ntawm cov kev ntsuas ntau yam ntawm kev sib txuas ua haujlwm, kev xav txog kev sib raug zoo (iCoh) tshwj xeeb tshaj yog kuaj pom "tseeb" kev sib cuam tshuam ntawm EEG cov cim tshwm sim nyob rau hauv lub sijhawm ncua sijhawm, yog li tsis quav ntsej kev cuam tshuam tam sim ntawm cov hluav taws xob nyob sib ze yuav tsim los ntawm qhov ntim ntawm cov hluav taws xob ua haujlwm los ntawm lub hlwb (Bullmore & Sporns, 2009; Fallani et al., 2014; Mohammad-Rezadeh li al., 2016).

iCoh qhov tseem ceeb tau suav rau txhua qhov ua tau electrodepairs hauv cov zaus hauv qab no: delta (1–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), spindle (12–15 Hz), beta (16– 30 Hz), thiab qis gamma (40-50 Hz).

Cov kev txheeb xyuas no tau ua tiav ob qho tib si Active thiab Sham cov ntaub ntawv. Tsis tas li ntawd, iCoh qhov tseem ceeb tau suav rau ob qho tib si poststimulation thiab prestimulation lub sij hawm qhov rais, thiab lub qhov rais prestimulation qhov tseem ceeb raug rho tawm los ntawm lub qhov rais poststimulation qhov tseem ceeb rau kev kho qhov hauv paus.Qhov siab multidimensionality ntawm iCoh ntsuas tau disentangled nrog ib daim duab theoreticapproach.

Daim Duab Ntsej Muag Inorastic muab cov ncauj lus rau cov pawg ntawm cov network uas muaj nqis rau kev sib txuas lus thiab kev sib koom tes ua ke los ua kev sib txuas lus thiab hloov kho cov ntaub ntawv ntawm lub hlwb cov cheeb tsam; Bullmore & Sporns, 2009;

Cov nram qab no feem ntau siv daim duab ntsuas tau xam rau tag nrho cov saum toj no-hais zaus bands nyob rau hauv lub pre-thiab poststimulation lub sij hawm: nruab nrab clustering coefficient (qhov tshwm sim ntawm cov nyob sib ze nodes raug kev cob cog rua, reflecting lub zos kev sib txuas); ntiaj teb no efficiency (yuav ua li cas npaum li cas lub network nyob rau hauv cov ntaub ntawv hloov lwm lub tsev); Cov yam ntxwv txoj kev ntev, lub vojvoog, thiab txoj kab uas hla (qhov nruab nrab ntawm cov npoo raws txoj kev luv tshaj plaws, qhov tsawg kawg nkaus qhov kev ncua deb, thiab qhov loj tshaj plaws nyob deb, nruab nrab ntawm txhua qhov ua tau ntawm cov nodes, ntsig txog); modularity (qhov degree uas lub paj hlwb sib cais rau hauv subnetworks los yog modules); assortativity (qhov kev faib ua feem ntawm cov nodes uas txuas rau lwm cov nodes uas zoo sib xws degrees vs.dissimilar degrees); ceev (tus naj npawb ntawm cov npoo muab faib los ntawm tus naj npawb ntawm cov nodes hauv daim duab); thiab txhais tau tias coherence (qhov nruab nrab coherence ntawm txhua khub ntawm nodes).

Classification thiab Regression Analyzes

Cov ntaub ntawv tshwm sim muaj 48 qhov kev sib txuas ua haujlwm zoo (8graph-theoretic metrics × 6 zaus bands) rau ob qho tib si Active thiab Sham stimulation tej yam kev mob rau txhua tus neeg koom.Qhov tshwj xeeb tshaj yog, kev sib txuas cov yam ntxwv tau muab rho tawm tom qab stimulation ntawm Active nightsin uas STAMP stimulation tshwm sim, thiab. raws li thaum lub sij hawm Sham hmo uas tsis muaj stimulation tshwm sim.

Txhawm rau tshawb xyuas qhov kev hloov pauv ntawm kev sib txuas yog kwv yees ntawm qhov kev xav tau, nrog rau kev cuam tshuam rau tus cwj pwm, peb tau ua raws li cov neeg koom nrog qib kev faib tawm thiab kev ntsuas kev rov ua dua tshiab, uas cov ntaub ntawv los ntawm ib pawg ntawm cov neeg koom tau raug tso tawm raws li cov ntaub ntawv ntsuas kev kwv yees. Txhawm rau txo qhov sib txawv, peb tau khiav 35 folds uas ib mus rau peb tus neeg koom tau tawm hauv ib lub sijhawm.

Qhov kev rov ua piv txwv ntawm qhov kev ua piv txwv no yog qhov zoo dua li qhov feem ntau siv "tso tawm" txoj kev, uas tuaj yeem ua rau tsis ruaj khov thiab tsis ncaj ncees rau kev kwv yees ntawm qhov sib txawv (Scheinost li al., 2019; Varoquaux li al., 2017). Peb tau xaiv los coj cov ntaub ntawv no-tsav approach tsis yog vim qhov loj ntawm qhov chaw ntawm cov ntaub ntawv tab sis kuj yog vim qhov tshiab ntawm kev tshawb nrhiav.

Ob peb cov kev tshawb fawb txog niaj hnub no tau tshuaj xyuas kev sib raug zoo ntawm cov duab-theoretic metrics ntawm lub hlwb kev sib txuas thiab metamemory rhiab heev, thiab tsis muaj leej twg tau tshuaj xyuas qhov kev hloov pauv no nrog lub hlwb stimulation. Yog li, peb yuav muaj peev xwm tsis nco qab kev sib raug zoo ntawm kev sib txuas ua haujlwm thiab metamemory los ntawm tsuas yog tshuaj xyuas qee zaus bands (xws li, theta) lossis ntsuas (xws li, txhais tau tias sib koom ua ke) raws li kev tshawb pom yav dhau los hauv cov ntaub ntawv. 2.

Rau txhua qhov quav, ib txoj kev xaiv xaiv tau khiav ntawm cov ntaub ntawv kev cob qhia kom pom cov duab-theoretic nta (poststimulationbaseline kho nrog prestimulation) uas muaj kev ntxub ntxaug ntawm Active thiab Sham stimulationconditions. Ntau yam nta tau muaj kev sib raug zoo nrog ib leeg, uas tuaj yeem pom hauv Cov Ntaub Ntawv Txhawb Nqa S2.

Yog li, thawj qhov sib txawv sib txawv heev tau raug tshem tawm los ntawm kev nrhiav cov khub ntawm cov yam ntxwv nrog kev sib raug zoo (threshold r tus nqi ntawm 0.8) thiab tshem tawm qhov sib txawv nrog qhov siab dua qhov kev sib raug zoo nrog rau lwm qhov sib txawv. Cov ntaub ntawv ntxiv ntawm kev tshem tawm cov hloov pauv vim muaj kev sib koom siab siab tuaj yeem pom nyob rau hauv Cov Ntaub Ntawv Txhawb Nqa.

Tom qab ntawd, ib qho kev ntsuam xyuas nyob ntawm (raws li txoj kev tshawb no yog ib qho kev tsim hauv cov ntsiab lus) tau khiav ntawm txhua qhov tshwj xeeb, sim qhov sib txawv ntawm Active thiab Sham tej yam kev mob. Cov kev hloov pauv uas tsis tshua muaj qhov tseem ceeb t (qhov pib ntawm 1) raug tshem tawm, nrog rau qhov laj thawj uas qhov sib txawv ntawm cov kev hloov pauv no tsawg heev.Cov yam ntxwv tseem ceeb tau nkag mus rau hauv Boruta feature xaiv algorithm (Kursa & Rudnicki, 2010).

Boruta yog ib txoj kev xaiv raws li permutation, nyob rau hauv uas txhua qhov feature, nrog rau ib daim qauv ntawm cov yam ntxwv nrog nws cov txiaj ntsig, shuffled hla cov neeg koom (hu ua "shadowfeature"), tau nkag mus rau hauv kev faib hav zoov random, thiab qhov tseem ceeb ntawm txhua qhov feature. yog txiav txim raws li tus qhab nias Z. Qhov tseem ceeb ntawm qhov tseeb nta yuav tsum siab tshaj qhov max Z qhab nia ntawm tus duab ntxoov ntxoo nta yuav suav hais tias tseem ceeb heev.

Cov txheej txheem no rov ua dua ob peb zaug nrog kev sib txawv ntawm cov duab ntxoov ntxoo kom tau txais kev faib tawm ntawm qhov tseem ceeb uas qhov tseem ceeb ntawm cov yam ntxwv tiag tiag tuaj yeem sib piv. Yog li, Boruta yog qhov tseem ceeb ntawm kev xaiv qhov tseem ceeb, txhua yam lossis tsis muaj qhov hloov pauv tuaj yeem raug xaiv los ntawm tus txheej txheem. Tsuas yog cov yam ntxwv uas muaj cov qhab nia Z siab dua li cov duab ntxoov ntxoo tau xaiv.

Cov algorithm no tau pom tias yuav txhim kho kev ua tau zoo hauv kev nrhiav cov ntsiab lus piv rau lwm qhov kev xaiv cov algorithms (Degenhardt li al., 2017; Kursa, 2017). Cov yam ntxwv uas tau xaiv los ntawm Boruta tau nkag mus rau hauv cov qauv logistic regression kwv yees lub Active tiv thaiv tej yam kev mob. Peb kuj tau khiav ib qho kev sib tw logistic tsis tu ncua nrog L2 tsis tu ncua (ridge regression), nyob rau hauv uas qhov kev pom zoo λ parameter tau txiav txim siab los ntawm cross-validation nyob rau hauv cov ntaub ntawv kawm.

Thaum kawg, peb kuj tau khiav qhov kev faib tawm logistic regression siv tsuas yog qhov tseem ceeb tshaj plaws hloov pauv tau xaiv los ntawm Boruta algorithm. Tom qab ua kom haum cov qauv ntsuas ntawm cov ntaub ntawv kev cob qhia, cov qauv tau siv los kwv yees cov ntaub ntawv xeem tawm sab laug, muab kev kwv yees ntawm qhov yuav tshwm sim rau qhov kev xeem piv txwv xws li teej tug mus rau Active lossis Sham mob. Cov kev tshwm sim no tau sib txuas ua ke thoob plaws folds thiab siv los xam cov txheeb ze kev ua haujlwm ntawm tus yam ntxwv nkhaus, los ntawm qhov peb tau muab los ntawm thaj chaw hauv qab qhov nkhaus (AUC) qhov tseem ceeb (Fawcett, 2006).

AUC yog qhov kev ntsuas ntau dua ntawm kev ua tau zoo ntawm kev ua haujlwm ntau dua li qhov tseeb ntawm qhov tseeb thiab yog li tau tshaj tawm ntawm no.Nyob rau hauv tib qhov kev lees paub kev sib tw, cov ntaub ntawv kev cob qhia tau siv los kwv yees cov kev hloov pauv hauv kev nco txog rau lub hom phiaj vignettes (xws li, Tag & Cue) los ntawm Hmo Ntuj 2 presleeptest rau qhov kev ntsuam xyuas postsleep nyob rau hauv lub Active mob.

Kev ntsuas ntawm tus cwj pwm kev ua tau zoo yog AUC ntawm hom -2 txheeb ze kev khiav hauj lwm tus yam ntxwv nkhaus, ib tug tsis ncaj ncees lawm nonparametric ntsuas ntawm metacognitive rhiab heev, nrog AUC ntau dua reflecting lub nco rhiab heev (Fleming &Lau, 2014; Galvin li al., 2003) . Ntawm no, cov qauv kwv yees qhov kev hloov thaum hmo ntuj hauv AUC los khi kev hloov pauv hauv kev sib txuas nrog kev hloov pauv hauv kev nco. Tag nrho cov connectivityfeatures tau nkag mus rau hauv ib tug zoo xws li feature xaiv txheej txheem raws li cov kev faib kev tsom xam.Highly collinear nta raug tshem tawm raws li ua ntej.

Hloov chaw ntawm qhov kev xeem, qhov kev sib raug zoo ntawm txhua qhov tshwj xeeb thiab kev hloov pauv AUC thaum hmo ntuj tau suav thiab cov yam ntxwv nrog tus nqi qis (qhov pib ntawm 0.1) raug tshem tawm. Cov yam ntxwv tseem ceeb tau khiav los ntawm Borutaalgorithm, lub sijhawm no kwv yees kev hloov pauv AUC hmo ntuj.

Cov yam ntxwv no, nrog rau tsuas yog qhov zoo tshaj plaws, tau nkag mus rau hauv cov kab rov tav kab rov tav thiab L2 cov qauv kev hloov pauv tsis tu ncua kwv yees cov ntaub ntawv xeem tawm sab laug. Lub hauv paus-mean-square yuam kev (RMSE) qhov tseem ceeb tau suav, thiab qhov nruab nrab RMSE qhov tseem ceeb thoob plaws folds tau muab piv rau RMSE qhov tseem ceeb los ntawm kev kwv yees los ntawm cov kab rov tav nrog tsuas yog cuam tshuam.

Raws li kev txheeb xyuas zaum kawg, peb tau coj Boruta-xaiv cov yam ntxwv los ntawm kev txheeb xyuas kev faib tawm, nrog rau Boruta-xaiv cov yam ntxwv los ntawm kev txheeb xyuas kev rov qab, thiab kev faib tawm rov qab thiab rov ua dua nrog cov txheej txheem ua ke, nrog rau cov yam ntxwv saum toj kawg nkaus ntawm txhua pawg. AUC qhov tseem ceeb los ntawm kev txheeb xyuas kev faib tawm thiab RMSE qhov tseem ceeb los ntawm kev txheeb xyuas kev rov qab tau tshaj tawm rau txhua qhov teeb tsa.

For more information:1950477648nn@gmail.com