混合尖峰發(fā)放模型-神經(jīng)科學(xué)課件

混合尖峰發(fā)放模型

混合尖峰發(fā)放模型

1Tounderstandhowthebrainworks,weneedtocombineexperimentalstudiesofanimalandhumannervoussystemswithnumericalsimulationoflarge-scalebrainmodels.

——E.M.IzhikevichTounderstandhowthebra2神經(jīng)元是生物體中信息傳遞的基本單元。神經(jīng)元是生物體中信息傳遞的基本單元。3文章解讀（Hybridspikingmodels）——severalkindsofmodelssimulationmethodsofsimplemodelofspiking neurons問題探究——重現(xiàn)圖2中不同神經(jīng)元的活動(dòng)模式，并指出 相應(yīng)活動(dòng)模式的參數(shù)統(tǒng)計(jì)各種模式下發(fā)放率與輸入刺激的強(qiáng)度之 間的關(guān)系輸入均值為符合泊松分布的隨機(jī)電脈沖， 考察均值與發(fā)放率之間關(guān)系文章解讀（Hybridspikingmodels）4Theleakyintegrate-and-firemodelv:membranepotentialC:memberanecapacitancegleak:leakyohmicconductanceEleak:leakyreversepotentialI:inputcurrentItlacksanintrinsicspike-generationmechanism,andhenceisjusta‘threshold’model.Theleakyintegrate-and-firem5Thequadraticintegrate-and-firemodelv:membranepotentialC:memberanecapacitanceI:inputcurrentK:apositiveparameterVrest:restingpotentialsVthresh:instantaneous thresholdpotentialsAhybridspikingmodel—asitcombinesasmoothspike-generationmechanismwithautocatalyticupstrokeofthespikeandahard‘a(chǎn)fter-spike’reset.Thequadraticintegrate-and-fi6SimplemodelofspikingneuronsU:arecoveryvariablea:thetimescaleofub:thesensitivityofuc:theafter-spikeresetvalueofvd:theafter-spikeresetofuSimplemodelofspikingneuron7SimplemodelofspikingneuronsU:arecoveryvariablea:thetimescaleofub:thesensitivityofuc:theafter-spikeresetvalueofvd:theafter-spikeresetofuSimplemodelofspikingneuron8OtherhybridspikingmodelsAllafter-spikingresettingarethesame.U：aconductanceE:thereversepotentialofuOtherhybridspikingmodelsAl9SimplemodelofspikingneuronsSimulationmethods1)HowtodetectVpeakcrossing?2)Strongsynapticconductancescausenumericalinstability.Howtoavoidnumericalinstability?Simplemodelofspikingneuron10SimplemodelofspikingneuronsSimulationmethodsDetectingVpeakcrossing:linearinterpolationSimplemodelofspikingneuron11Avoidingnumericalinstability:SimplemodelofspikingneuronsSimulationmethodstime-varyingconductancethereversepotentialforaparticularsynapticcurrentie.g.i=NMDA,AMPA,GABAAandGABABAvoidingnumericalinstability12Avoidingnumericalinstability:SimplemodelofspikingneuronsSimulationmethodsThetotalconductanceThetotalreversepotentialAvoidingnumericalinstability13Avoidingnumericalinstability:SimplemodelofspikingneuronsSimulationmethodsIndeed,positivevaluesofg(t)pushthemembranevoltagetowardsthereversepotentialE(t).However,wheng(t)islarge,thetermtg(t)[E(t)?v(t)]becomeslarge,resultinginovershootanddivergencefromE(t).Noticethat,nomatterhowsmallthesimulationsteptis,theconductancevariablecouldalwaysbecomesolargeastocreatetheinstability.Avoidingnumericalinstability14Avoidingnumericalinstability:SimplemodelofspikingneuronsSimulationmethodsFinally，Sincethenumericalinstabilityiscausedbythelinearterm,onecanusethehybridnumericalmethodthatcombinesthesimplicityandefficiencyofexplicitmethodsandthenumericalstabilityofimplicitmethods.Avoidingnumericalinstability15Avoidingnumericalinstability:SimplemodelofspikingneuronsSimulationmethodsLeft:Right:Avoidingnumericalinstability16問題探究一

重現(xiàn)圖2中不同神經(jīng)元的活動(dòng)模式，并指出相應(yīng)活動(dòng)模式的參數(shù)——form“Simplemodelofspikingneurons”

問題探究一重現(xiàn)圖2中不同神經(jīng)元的活動(dòng)模式，并指出相應(yīng)17問題探究一:程序示例subplot(5,4,1)a=0.02;b=0.2;c=-65;d=6;V=-70;u=b*V;VV=[];uu=[];tau=0.25;tspan=0:tau:100;T1=tspan(end)/10;fort=tspanif(t>T1)I=14;elseI=0;end;V=V+tau*(0.04*V^2+5*V+140-u+I);u=u+tau*a*(b*V-u);ifV>30VV(end+1)=30;V=c;u=u+d;elseVV(end+1)=V;end;uu(end+1)=u;end;plot(tspan,VV,[0T1T1max(tspan)],-90+[001010]);axis([0max(tspan)-9030])axisoff;title('(A)tonicspiking');問題探究一:程序示例subplot(5,4,1)ifV18問題探究一:figure1問題探究一:figure119問題探究一:figure2問題探究一:figure220問題探究一:參數(shù)問題探究一:參數(shù)21問題探究一:參數(shù)問題探究一:參數(shù)22問題探究二統(tǒng)計(jì)各種模式下發(fā)放率和輸入刺激的強(qiáng)度之間的關(guān)系1）在不同模式下，改變輸入刺激電流I的大小，并統(tǒng)計(jì)不同的輸入電流I對應(yīng)的發(fā)放率r，利用plot函數(shù)繪制不同模式下的“r-I”關(guān)系圖2）對A模式（tonicspiking）的“r-I”關(guān)系圖進(jìn)行擬合，得出其對應(yīng)表達(dá)式注：統(tǒng)計(jì)發(fā)放率方法眾多，此處采用較為簡便的“時(shí)間平均法”問題探究二統(tǒng)計(jì)各種模式下發(fā)放率和輸入刺激的強(qiáng)度之間的關(guān)系1）23問題探究二：程序示例V=V+tau*(0.04*V^2+5*V+140-u+I);u=u+tau*a*(b*V-u);ifV>30VV(end+1)=30;V=c;u=u+d;elseVV(end+1)=V;end;uu(end+1)=u;end;ffl(i)=histc(VV,30)/100;endfigureplot(dl,ffl,'-o')title('(A)tonicspiking');subplot(5,4,1)a=0.02;b=0.2;c=-65;d=6;V=-70;u=b*V;VV=[];uu=[];tau=0.25;tspan=0:tau:100;T1=tspan(end)/10;dl=1:50;ffl=zeros(1,length(dl));fori=1:length(dl)fort=tspanif(t>T1)I=dl(i);elseI=0;end;問題探究二：程序示例V=V+tau*(0.04*V^24問題探究二：“r-I”關(guān)系圖問題探究二：“r-I”關(guān)系圖25問題探究二對A模式（tonicspiking）的“r-I”關(guān)系圖進(jìn)行擬合LinearmodelPoly2:f(x)=p1*x^2+p2*x+p3Coefficients(with95%confidencebounds):p1=0.001364(0.001358,0.001369)p2=6.425e-005(-0.0002163,0.0003448)p3=-0.007935(-0.01104,-0.004833)即：問題探究二對A模式（tonicspiking）的“r-I”26問題探究三輸入均值為符合泊松分布的隨機(jī)電脈沖，考察均值和發(fā)放率之間的關(guān)系(1)選擇“tonicspiking”的脈沖為基本脈沖單元，在一個(gè)單位的時(shí)間內(nèi)有n