본문

서브메뉴

Gait Signatures: Data-Driven Discovery of Individual-Specific Neuromechanical Dynamics
Gait Signatures: Data-Driven Discovery of Individual-Specific Neuromechanical Dynamics
Gait Signatures: Data-Driven Discovery of Individual-Specific Neuromechanical Dynamics

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20260202105549
ISBN  
9798265400178
DDC  
612.8
저자명  
Winner, Taniel S.
서명/저자  
Gait Signatures: Data-Driven Discovery of Individual-Specific Neuromechanical Dynamics
발행사항  
[Sl] : Georgia Institute of Technology, 2024
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2024
형태사항  
228 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-05, Section: B.
주기사항  
Advisor: Ting, Lena.
학위논문주기  
Thesis (Ph.D.)--Georgia Institute of Technology, 2024.
초록/해제  
요약Human gait is important because it is fundamental to the execution of daily activities which impacts quality of life. Notably, changes to individuals' gait patterns can indicate underlying health issues such as musculoskeletal and neurological disorders, underscoring the importance of gait analysis towards rehabilitation following disease, injury and surgery. However, standard methods often focus on discrete gait variables, neglecting continuous data and complex inter-limb and inter-joint spatiotemporal dependencies that underly gait and impairment. A comprehensive approach is necessary to analyze gait patterns, diagnose conditions accurately, prescribe treatments, develop tailored rehabilitation strategies and monitor progress effectively.This thesis introduces a novel, data-driven approach to model the dynamics of human gait, effectively capturing the complex spatiotemporal dependencies between individuals' joint angles, arising from joint neural and biomechanical constraints. Individual-specific representations of gait dynamics, termed "gait signatures", offer a holistic way to identify and quantify individual-specific differences in gait in both health and disease. I use continuous gait kinematics from multiple individuals to train a single recurrent neural network (RNN) to predict the time evolution of multi-joint gait coordination, thereby capturing the underlying dynamics of human gait. By recapitulating the gait behaviors of multiple individuals into common low-dimensional basis functions, this approach facilitates holistic comparison of gait dynamics across different individuals and cohorts in a shared gait space. This method is particularly valuable for modeling complex gait deviations, including those seen in individuals with neurological disorders, as it eliminates the challenging task of estimating physiological constraints like neural control.In a series of studies, I demonstrate that highly individualized gait dynamics (i.e. gait signatures) are conserved across walking speed for both able-bodied (AB) adults and stroke survivors during treadmill walking. Notably, AB gait signatures show predictable, linear changes across a wide range of speeds. These gait signatures, comprising of biomechanically interpretable subcomponents, can be manipulated to understand their relationship with observed joint coordination patterns. Additionally, using the RNN gait dynamics model, I demonstrate its ability to predict the time evolution of multi-joint coordination from an initial posture. Finally, I utilized the gait signatures framework to track and quantify clinically meaningful changes in stroke survivors' gait signatures through two longitudinal gait interventions (Fast and Fast Functional Electrical Stimulation). These findings underscore the potential of gait signatures for evaluating and interpreting treatment induced changes in gait in a holistic manner.This work demonstrates promising clinical applications in precision medicine. Gait signatures offer an objective, reliable and repeatable method to differentiate and analyze individuals' unique dynamics regardless of walking speed. They can predict the potential response to gait interventions and perturbations, facilitating the development of personalized strategies to improve gait while reducing time and participant burden.
일반주제명  
Brain research
일반주제명  
Tendons
일반주제명  
Electromyography
일반주제명  
Dynamical systems
일반주제명  
Ataxia
일반주제명  
Human subjects
일반주제명  
Gait
일반주제명  
Stroke
일반주제명  
Partial differential equations
일반주제명  
Musculoskeletal system
일반주제명  
Neurological disorders
일반주제명  
Neural networks
일반주제명  
Support vector machines
일반주제명  
Kinetics
일반주제명  
Nervous system
일반주제명  
Muscle function
일반주제명  
Posture
일반주제명  
Biomechanics
일반주제명  
Kinesiology
일반주제명  
Computer science
일반주제명  
Mathematics
일반주제명  
Medicine
일반주제명  
Morphology
일반주제명  
Neurosciences
기타저자  
Georgia Institute of Technology.
기본자료저록  
Dissertations Abstracts International. 87-05B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2024        us                              c    eng  d
■001000017360577
■00520260202105549
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798265400178
■035    ▼a(MiAaPQ)AAI32315671
■035    ▼a(MiAaPQ)GeorgiaTech75674
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a612.8
■1001  ▼aWinner,  Taniel  S.
■24510▼aGait  Signatures:  Data-Driven  Discovery  of  Individual-Specific  Neuromechanical  Dynamics
■260    ▼a[Sl]▼bGeorgia  Institute  of  Technology▼c2024
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2024
■300    ▼a228  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-05,  Section:  B.
■500    ▼aAdvisor:  Ting,  Lena.
■5021  ▼aThesis  (Ph.D.)--Georgia  Institute  of  Technology,  2024.
■520    ▼aHuman  gait  is  important  because  it  is  fundamental  to  the  execution  of  daily  activities  which  impacts  quality  of  life.  Notably,  changes  to  individuals'  gait  patterns  can  indicate  underlying  health  issues  such  as  musculoskeletal  and  neurological  disorders,  underscoring  the  importance  of  gait  analysis  towards  rehabilitation  following  disease,  injury  and  surgery.  However,  standard  methods  often  focus  on  discrete  gait  variables,  neglecting  continuous  data  and  complex  inter-limb  and  inter-joint  spatiotemporal  dependencies  that  underly  gait  and  impairment.  A  comprehensive  approach  is  necessary  to  analyze  gait  patterns,  diagnose  conditions  accurately,  prescribe  treatments,  develop  tailored  rehabilitation  strategies  and  monitor  progress  effectively.This  thesis  introduces  a  novel,  data-driven  approach  to  model  the  dynamics  of  human  gait,  effectively  capturing  the  complex  spatiotemporal  dependencies  between  individuals'  joint  angles,  arising  from  joint  neural  and  biomechanical  constraints.  Individual-specific  representations  of  gait  dynamics,  termed  "gait  signatures",  offer  a  holistic  way  to  identify  and  quantify  individual-specific  differences  in  gait  in  both  health  and  disease.  I  use  continuous  gait  kinematics  from  multiple  individuals  to  train  a  single  recurrent  neural  network  (RNN)  to  predict  the  time  evolution  of  multi-joint  gait  coordination,  thereby  capturing  the  underlying  dynamics  of  human  gait.  By  recapitulating  the  gait  behaviors  of  multiple  individuals  into  common  low-dimensional  basis  functions,  this  approach  facilitates  holistic  comparison  of  gait  dynamics  across  different  individuals  and  cohorts  in  a  shared  gait  space.  This  method  is  particularly  valuable  for  modeling  complex  gait  deviations,  including  those  seen  in  individuals  with  neurological  disorders,  as  it  eliminates  the  challenging  task  of  estimating  physiological  constraints  like  neural  control.In  a  series  of  studies,  I  demonstrate  that  highly  individualized  gait  dynamics  (i.e.  gait  signatures)  are  conserved  across  walking  speed  for  both  able-bodied  (AB)  adults  and  stroke  survivors  during  treadmill  walking.  Notably,  AB  gait  signatures  show  predictable,  linear  changes  across  a  wide  range  of  speeds.  These  gait  signatures,  comprising  of  biomechanically  interpretable  subcomponents,  can  be  manipulated  to  understand  their  relationship  with  observed  joint  coordination  patterns.  Additionally,  using  the  RNN  gait  dynamics  model,  I  demonstrate  its  ability  to  predict  the  time  evolution  of  multi-joint  coordination  from  an  initial  posture.  Finally,  I  utilized  the  gait  signatures  framework  to  track  and  quantify  clinically  meaningful  changes  in  stroke  survivors'  gait  signatures  through  two  longitudinal  gait  interventions  (Fast  and  Fast  Functional  Electrical  Stimulation).  These  findings  underscore  the  potential  of  gait  signatures  for  evaluating  and  interpreting  treatment  induced  changes  in  gait  in  a  holistic  manner.This  work  demonstrates  promising  clinical  applications  in  precision  medicine.  Gait  signatures  offer  an  objective,  reliable  and  repeatable  method  to  differentiate  and  analyze  individuals'  unique  dynamics  regardless  of  walking  speed.  They  can  predict  the  potential  response  to  gait  interventions  and  perturbations,  facilitating  the  development  of  personalized  strategies  to  improve  gait  while  reducing  time  and  participant  burden.
■590    ▼aSchool  code:  0078.
■650  4▼aBrain  research
■650  4▼aTendons
■650  4▼aElectromyography
■650  4▼aDynamical  systems
■650  4▼aAtaxia
■650  4▼aHuman  subjects
■650  4▼aGait
■650  4▼aStroke
■650  4▼aPartial  differential  equations
■650  4▼aMusculoskeletal  system
■650  4▼aNeurological  disorders
■650  4▼aNeural  networks
■650  4▼aSupport  vector  machines
■650  4▼aKinetics
■650  4▼aNervous  system
■650  4▼aMuscle  function
■650  4▼aPosture
■650  4▼aBiomechanics
■650  4▼aKinesiology
■650  4▼aComputer  science
■650  4▼aMathematics
■650  4▼aMedicine
■650  4▼aMorphology
■650  4▼aNeurosciences
■690    ▼a0648
■690    ▼a0575
■690    ▼a0800
■690    ▼a0984
■690    ▼a0405
■690    ▼a0564
■690    ▼a0287
■690    ▼a0317
■71020▼aGeorgia  Institute  of  Technology.
■7730  ▼tDissertations  Abstracts  International▼g87-05B.
■790    ▼a0078
■791    ▼aPh.D.
■792    ▼a2024
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360577▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

미리보기

내보내기

chatGPT토론

Ai 추천 관련 도서


    신착도서 더보기
    최근 3년간 통계입니다.

    소장정보

    • 예약
    • 소재불명신고
    • 나의폴더
    • 우선정리요청
    • 비도서대출신청
    • 야간 도서대출신청
    소장자료
    등록번호 청구기호 소장처 대출가능여부 대출정보
    TF14959 전자도서 대출가능 마이폴더 부재도서신고 비도서대출신청 야간 도서대출신청

    * 대출중인 자료에 한하여 예약이 가능합니다. 예약을 원하시면 예약버튼을 클릭하십시오.

    해당 도서를 다른 이용자가 함께 대출한 도서

    관련 인기도서

    로그인 후 이용 가능합니다.