Grant: Master's Award: Canada Graduate Scholarships
McGill University
$17,500.00
Learning Sleep Stages from deep brain recordings to improve treatment methods in Parkinson's Disease
EF 2019-20
Résumé
"My proposed research project creates predictive algorithms for use in adaptive deep brain stimulation, a treatment for severe Parkinson's Disease.Parkinson's disease (PD), which primarily affects older individuals, is the second most common neurodegenerative disorder in Canada. Individuals with severe, drug-resistant PD are treated with deep brain stimulation (DBS), wherein an electrode is implanted within the brain to electrically stimulate nerve cells. DBS is effective in reducing symptoms, but could be made more so with adaptive systems that respond to patient state. However, patients cannot turn off systems themselves due to PD symptoms which it difficult for patients to independently turn the systems on. Thus automated systems are required, but algorithms which determine the patient state and how to change stimulation are necessary, as current techniques are rudimentary and may not be applicable to all people. My research goal is to develop such algorithms that can predict sleep stage from local field potentials recorded with the implanted electrode to fill this bridge.Two thrusts of my research are to investigate the effects of inter-patient variability on sleep stages in order to improve generalization of the nets, and to incorporate time information into the systems to create systems that make predictions about current and future states of the patient. To do so, I will use machine learning techniques such as Siamese networks, which can distill information specific to patient and states, to gain a better idea of how stages of sleep vary. I will also incorporate temporal information through filtering techniques and analysis improvements and develop methods to predict whether transitions into or out of sleep are imminent."
Programme
Master's Award: Canada Graduate Scholarships
Bénéficiaire
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
CIHR Funding Decisions Database
Tel que publié
Id
429091
Name
Saltoun Karin M
Region
Québec
Country
Canada
Orgtype
Institution Paid; Research Institution
Prcname2
Special Cases - Awards Programs
Projecttitle
Learning Sleep Stages from deep brain recordings to improve treatment methods in Parkinson's Disease
Programtype2
Studentship Programs
Orgnameinp2
McGill University
Cihramount2
$17,500
Cihrequipment2
$0
Programname2
Master's Award: Canada Graduate Scholarships
Keyworddelim
Computational Methods; Deep Brain Stimulation; Machine Learning; Parkinson'S Disease; Sleep
Abstract
"My proposed research project creates predictive algorithms for use in adaptive deep brain stimulation, a treatment for severe Parkinson's Disease.Parkinson's disease (PD), which primarily affects older individuals, is the second most common neurodegenerative disorder in Canada. Individuals with severe, drug-resistant PD are treated with deep brain stimulation (DBS), wherein an electrode is implanted within the brain to electrically stimulate nerve cells. DBS is effective in reducing symptoms, but could be made more so with adaptive systems that respond to patient state. However, patients cannot turn off systems themselves due to PD symptoms which it difficult for patients to independently turn the systems on. Thus automated systems are required, but algorithms which determine the patient state and how to change stimulation are necessary, as current techniques are rudimentary and may not be applicable to all people. My research goal is to develop such algorithms that can predict sleep stage from local field potentials recorded with the implanted electrode to fill this bridge.Two thrusts of my research are to investigate the effects of inter-patient variability on sleep stages in order to improve generalization of the nets, and to incorporate time information into the systems to create systems that make predictions about current and future states of the patient. To do so, I will use machine learning techniques such as Siamese networks, which can distill information specific to patient and states, to gain a better idea of how stages of sleep vary. I will also incorporate temporal information through filtering techniques and analysis improvements and develop methods to predict whether transitions into or out of sleep are imminent."
Competitiondate
201912
Approvedterm2
1 yr 0 mth
Pinamesdelim
Saltoun, Karin M
Approvedterm
1 an 0 mois
Cihrcontribution2
~ $17,500~ $0~