Table Of ContentLecture Notes in Mathematics 2058
Editors:
J.-M.Morel,Cachan
B.Teissier,Paris
EditorsMathematicalBiosciencesSubseries:
P.K.Maini,Oxford
Forfurthervolumes:
http://www.springer.com/series/304
•
Mostafa Bachar
Jerry Batzel
Susanne Ditlevsen
Editors
Stochastic
Biomathematical
Models
with Applications to Neuronal Modeling
123
Editors
MostafaBachar SusanneDitlevsen
KingSaudUniversity UniversityofCopenhagen
CollegeofSciences DepartmentofMathematicalSciences
DepartmentofMathematics Copenhagen,Denmark
Riyadh,SaudiArabia
JerryBatzel
UniversityofGraz
InstituteforMathematics
andScientificComputing
andMedicalUniversityofGraz
InstituteofPhysiology
Graz,Austria
ISBN978-3-642-32156-6 ISBN978-3-642-32157-3(eBook)
DOI10.1007/978-3-642-32157-3
SpringerHeidelbergNewYorkDordrechtLondon
LectureNotesinMathematicsISSNprintedition:0075-8434
ISSNelectronicedition:1617-9692
LibraryofCongressControlNumber:2012949578
MathematicsSubjectClassification(2010):60Gxx,92C20,37N25,92Bxx
(cid:2)c Springer-VerlagBerlinHeidelberg2013
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof
thematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation,
broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionorinformation
storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodology
nowknownorhereafterdeveloped.Exemptedfromthislegalreservationarebriefexcerptsinconnection
with reviews or scholarly analysis or material supplied specifically for the purpose of being entered
and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of
this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’slocation,initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.
PermissionsforusemaybeobtainedthroughRightsLinkattheCopyrightClearanceCenter.Violations
areliabletoprosecutionundertherespectiveCopyrightLaw.
Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication
doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant
protectivelawsandregulationsandthereforefreeforgeneraluse.
While the advice and information in this book are believed to be true and accurate at the date of
publication,neithertheauthorsnortheeditorsnorthepublishercanacceptanylegalresponsibilityfor
anyerrorsoromissionsthatmaybemade.Thepublishermakesnowarranty,expressorimplied,with
respecttothematerialcontainedherein.
Printedonacid-freepaper
SpringerispartofSpringerScience+BusinessMedia(www.springer.com)
Preface
Why use mathematical or stochastic models to study something as complicated
andoftenpoorlyunderstoodasdynamicsin physiology?Hopefully,thisbookcan
providesomepartialanswersandpointtotheexcitingproblemsthatstillremainto
besolved,wheremathematicalandstochastictoolscanbeuseful.
In this book we treat basics of stochastic process theory represented by a
stochastic differential equation directed towards biological modeling and review
the field of neuronalmodels. Theoreticalmodelsmust be relevantphysiologically
to be useful and interesting, and their analysis can provide biological insight
and help summarize and interpret experimentaldata. Predictions can be extracted
from the model, and experiments verifying or invalidating the model can be
suggested,therebyenhancingphysiologicalunderstanding.Evenifthemechanisms
are well understood, simulations from models can explore the consequences of
extremephysiologicalconditionsthatmightbeunethicalorimpossibletoreproduce
experimentally.Theprocessofbuildingatheoreticalmodelforcesonetoconsider
anddecideontheessentialcharacteristicsofthephysiologicaldynamics,aswellas
which variables and mechanisms to include. Analysis and numerical simulations
of the model illustrate quantitatively and qualitatively the consequences of the
assumptions implied in the model. The unifying aim of theoretical modeling and
experimentalinvestigationistheelucidationoftheunderlyingbiologicalprocesses
thatresultinaparticularobservedphenomenon.
Manybiologicalsystemsarehighlyirregular,andexperimentsundercontrolled
conditionsshowalargetrial-to-trialvariability,evenwhenkeepingtheexperimental
setup fixed. This calls for a stochastic, as opposed to deterministic, modeling
approach, especially because ignoring the stochastic phenomena in the modeling
mayhugelyaffecttheconclusionsofthestudiedbiologicalsystem.Inlinearsystems
the noise might only blur the underlyingdynamicswithout qualitatively affecting
it, but in nonlinear dynamical systems corrupted by noise, the corresponding
deterministic dynamics can be drastically changed. In general, stochastic effects
influence the dynamics and may enhance, diminish, or even completely change
thedynamicbehaviorofthesystem.Incertainbiologicalsystems,e.g.,inauditory
v
vi Preface
neurons,thenoiseisevenbelievedtoenhancethesignal,thusprovidingabiological
justificationforthelargeamountofnoisefoundinlivingsystems.
Thebooktreatsstochasticityrepresentedbystochasticdifferentialequations,but
isnotmeanttobeacomprehensivetextbookonstochasticmethods.Itisprimarily
intended for mathematicians and life scientists who are interested in seeing an
in-depth and motivated presentation of an important application of stochastic
methods. Our goal is to provide an illuminating example of where and how
stochastic methods can enter into modeling physiological systems. Life scientists
generally, with some background in mathematics, will also be able to benefit
by seeing how these two areas are linked, and we point the interested reader to
referencesthatfillinmissingbackgroundinformation.Wehopethatthematerialas
presentedwillprovideausefulillustrationofinterdisciplinaryresearch.
Thereadershouldhavea basicbackgroundinprobability,differentialandinte-
gralcalculus,andordinarydifferentialequations.Wefocusonneuronalmodeling,
wherestochasticmodelshavesupplementeddeterministicdynamicalmodelssince
the sixties. There exist already many excellent books on nonlinear dynamics,
biomathematicalmodeling, and computationalneuroscience. The aim of this vol-
ume is to provide a focused, up-to-date presentation of neuronal modeling using
stochastic methods, while in addition both motivating the linkage and providing
insightintopracticalissues.Moreprofitwillofcoursebederivedfromreadingthe
book,ifitiscombinedwiththereadingofsomeintroductorytextsincomputational
neuroscienceorbiomathematicalmodeling.
Thebookisdividedintotwoparts.Thefirstpartintroducessomemethodology,
which is useful when modeling biological systems with stochastic dynamics.
Chapter 1 is an introduction to stochastic models and a good place to start if the
readerhasnobackgroundinstochasticprocesses.Alsoashortoverviewofstatistical
methodstoestimatemodelparametersfromdataisprovided.Chapters2and3need
more mathematical preparation and a basic understanding of stochastic processes
andprobabilitytheory,aswellassomenotionofmeasuretheory.InChap.2,scale
andspeedmeasuresofone-dimensionaldiffusionprocessesarereviewed,whichare
used to determineexpressionsfor hitting times or first-escape times, in particular,
boundarybehaviors.Thesetoolsareveryusefulwhendealingwithneuronalmodels,
as thosepresentedin the secondpartof the book.Chapter3 givesan introduction
to the theory of large deviations, providing an asymptotic description of the
fluctuationsofastochasticsystem,inparticular,providingexponentialestimatesfor
thewaitingtimetorareevents.Chapter4closesthemethodologicalpartindicating
howthetheoryfromChap.3canbeimplementedinpractice.Thesetechniquesare
naturaltousewhenanalyzingstochasticmodelsofbiologicalsystems.
Thesecondpartofthebooktreatsneuronalmodels.Chapter5providesatimely
review of existing methods and available analytical results for the most common
one-dimensional stochastic integrate-and-fire models. These models of neuronal
activity collapse the neuronal anatomy into a single point in space, sacrificing
realism for mathematical tractability, although they often succeed in predicting
neuronalresponsewithconsiderableaccuracy.Chapter6goesastepfurthertomore
realisticmodelsandincludesthespatialdimensionofneuronaldynamics.Stochastic
Preface vii
partialdifferentialequationmodelsarereviewed,inparticulartheHodgkin–Huxley
andtheFitzHugh–Nagumomodelsaretreatedindetail.Chapter7isdedicatedtoa
probabilistictreatmentofFitzHugh–Nagumosystems.Finally,Chap.8implements
the tools from Chap.6 to a specific application of modeling spreading of cortical
depression.
This volume was first conceivedas a result of the experienceof designing and
holding a combined summer school/workshop event on the subject of stochas-
tic modeling in physiology. This event was part of a Marie Curie sponsored
series of four training events designed to encourage interdisciplinary research
in modeling physiological systems. The events brought together mathematicians,
bioengineers, statisticians, medical clinicians, physiologists, and other related life
scienceresearchers.
Theunderlyingmotivationandinspirationforthisseriesofeventsisthatunrav-
eling the complexities of physiological systems and the intricacies of interaction
between systems requires development of novel and innovative insights as well
as new research approaches and techniques. Furthermore, such new approaches
can be strongly stimulated by merging the different perspectives from the mathe-
matical/engineeringdisciplines on the one side and the life sciences on the other.
These observations motivated the design of the events in which summer schools
wouldintroducenewandyoungresearchestoaninterdisciplinarytreatmentofthe
modelingofkeyphysiologicalsystemswithemphasisonhowmodelingcanaddress
important clinical problems related to these systems. Directly following each
summerschoolaninterdisciplinaryscientificworkshopwasheld.Theseworkshops
focused on the same themes as the preceding summer school and were designed
as stand alone scientific events. Students from the school participated in these
workshopsandinthiswaynewandcurrentresearcherscouldinteractanddevelop
contactsforcollaboration.
ThegeneralwebpagelinkingandreflectingallfourMarieCurietrainingeventscan
befoundat:http://www.uni-graz.at/biomedmath/info.html.
Theeventrelatedtothisvolumecanbefoundattheeventwebpage:
http://www.math.ku.dk/(cid:2)susanne/SummerSchool2008/
Acknowledgements This book would never have existed without the work of the contributors
ofthisvolume.Theyallparticipatedinandcontributedtothesuccessofthecombinedsummer
school and workshop, and we would like to thank them all for sharing their insights with us.
WearegratefultoMichaelSørensenandFranzKappelfortheirhelpandpartoforganizingthe
summerschoolandtheworkshop.Wewishtothankallthepeoplewhoattendedthesummerschool
and workshop and especially theteachers of thesummer school: Andrea DeGaetano, Susanne
Ditlevsen,TereseGraversen,MartinJacobsen,SeemaNanda,BerntØksendal,UmbertoPicchini,
Laura Sacerdote, Michael Sørensen, and Gilles Wainrib. Thanks to Flemming H. Jacobsen for
typing parts of the book. We also thank Springer Verlag and especially Ute McCrory for their
support during the production of the book. Finally, this volume would not have been possible
withoutthefunds from different sources, mainlytheEuropean Unionundertheprogram Marie
CurieConferencesandTrainingCoursewhichsupportedthesummerschoolandworkshopunder
viii Preface
theprojectBiomathtech07-10(MSCF-CT-2006-045961).WewerealsosupportedbytheEuropean
Society for Mathematical and Theoretical Biology, Forskerskolen i Matematik og Anvendelser,
Forskerskolen i Biostatistik, Copenhagen Statistics Network at University of Copenhagen and
Biomedical Simulations Resource, University of Southern California. Research by Susanne
DitlevsenwassupportedbytheDanishCouncilforIndependentResearchjNaturalSciences,Jerry
BatzelwaspartiallyfundedbyFWF(Austria)projectP18778-N13,MostafaBacharwaspartially
supported by Deanship of Scientific Research, College of Science Research center, King Saud
University,AdelineSamsonwassupportedbytheBonusQualiteRecherchefromUniversiteParis
Descartes, Laura Sacerdote and Maria Teresa Giraudo were supported by MIUR PRIN 2008,
Miche`leThieullenwasSupportedbyAgenceNationaledeRecherche ANR-09-BLAN-0008-01,
andHenryTuckwellthanksProfDrJu¨rgenJostforhisfinehospitalityatMISMPI.
Riyadh,SaudiArabia MostafaBachar
Graz,Austria JerryBatzel
Copenhagen,Denmark SusanneDitlevsen
February2012
Contents
PartI Methodology
1 IntroductiontoStochasticModelsinBiology............................. 3
SusanneDitlevsenandAdelineSamson
1.1 Introduction ............................................................. 3
1.2 MarkovChainsandDiscrete-TimeProcesses ......................... 4
1.3 TheWienerProcess(orBrownianMotion)............................ 5
1.4 StochasticDifferentialEquations ...................................... 8
1.5 ExistenceandUniqueness.............................................. 13
1.6 Itoˆ’sFormula............................................................ 14
1.7 MonteCarloSimulations............................................... 16
1.7.1 TheEuler–MaruyamaScheme................................ 16
1.7.2 TheMilsteinScheme.......................................... 17
1.8 Inference................................................................. 17
1.8.1 MaximumLikelihood......................................... 18
1.8.2 BayesianApproach............................................ 22
1.8.3 MartingaleEstimatingFunctions............................. 23
1.9 BiologicalApplications................................................. 25
1.9.1 Oncology....................................................... 25
1.9.2 Agronomy ..................................................... 29
References..................................................................... 33
2 One-DimensionalHomogeneousDiffusions............................... 37
MartinJacobsen
2.1 Introduction ............................................................. 37
2.2 DiffusionProcesses..................................................... 39
2.3 ScaleFunctionandSpeedMeasure .................................... 40
2.4 BoundaryBehavior ..................................................... 45
2.5 ExpectedTimetoHitaGivenLevel................................... 54
References..................................................................... 55
ix
x Contents
3 ABriefIntroductiontoLargeDeviationsTheory........................ 57
GillesWainrib
3.1 Introduction ............................................................. 57
3.2 SumofIndependentRandomVariables................................ 58
3.3 GeneralTheory.......................................................... 61
3.4 SomeLargeDeviationsPrinciplesforStochasticProcesses.......... 64
3.4.1 SanovTheoremforMarkovChains.......................... 64
3.4.2 SmallNoiseandFreidlin–WentzellTheory.................. 65
3.5 Conclusion .............................................................. 71
References..................................................................... 71
4 SomeNumericalMethodsforRareEventsSimulationand
Analysis ....................................................................... 73
GillesWainrib
4.1 Introduction ............................................................. 73
4.2 Monte-CarloSimulationMethods...................................... 75
4.2.1 OverviewoftheDifferentApproaches....................... 75
4.2.2 FocusonImportanceSampling............................... 78
4.3 NumericalMethodsBasedonLargeDeviationsTheory.............. 87
4.3.1 QuasipotentialandOptimalPath ............................. 87
4.3.2 NumericalMethods ........................................... 88
4.4 Conclusion .............................................................. 93
References..................................................................... 95
PartII NeuronalModels
5 Stochastic Integrate and Fire Models: A Review on
MathematicalMethodsandTheirApplications.......................... 99
LauraSacerdoteandMariaTeresaGiraudo
5.1 Introduction ............................................................. 99
5.2 BiologicalFeaturesoftheNeuron ..................................... 101
5.3 OneDimensionalStochasticIntegrateandFireModels .............. 102
5.3.1 IntroductionandNotation..................................... 102
5.3.2 WienerProcessModel ........................................ 103
5.3.3 RandomizedRandomWalkModel........................... 105
5.3.4 Stein’sModel.................................................. 106
5.3.5 Ornstein–UhlenbeckDiffusionModel ....................... 107
5.3.6 ReversalPotentialModels .................................... 112
5.3.7 ComparisonBetweenDifferentLIFModels................. 117
5.3.8 JumpDiffusionModels....................................... 119
5.3.9 BoundaryShapes.............................................. 120
5.3.10 FurtherModels ................................................ 121
5.3.11 RefractorinessandReturnProcessModels .................. 122
