Grant: Master's Award: Canada Graduate Scholarships
University of Ottawa
$27,000.00
Using Machine Learning and Precision Medicine to Identify Youth at Risk of Relapsing Eating Disorders
EF 2024-25
Résumé
Background:Eating disorders (ED) are serious mental health conditions with high mortality and relapse rates, often emerging during adolescence. Early intervention with personalized treatment could improve outcomes, but the inability to accurately predict disease course and relapse risk hinders this potential. Application of machine learning (ML) and precision medicine (PM) biological investigations has been proposed as promising strategies to overcome these limitations and enhance ED relapse prediction.Objectives:1.
Assess if ML improves prediction of relapsing ED in adolescents using intake assessment data.2.
Evaluate the added value of PM investigations in predicting relapsing ED.3.
Identify key factors contributing to relapse risk across biopsychosocial domains.Methods:Longitudinal data (2018-2024) from 383 patients in the CHEO ED program, including clinical, biological, psychological, and social metrics collected at intake across all treatment levels will be analyzed. A subset of patients (n=95) will also have PM data including microbiome, inflammatory, endocrine, sleep, smell, and taste testing. Predictive modeling of relapsing ED status will be compared between ML methods (Random Forest and Elastic Net) and logistic regression. Models will be built with and without PM biological data to assess the utility of multi-axial data.Significance:This will be the first study to apply both ML and PM to enhance prediction and understanding of ED relapse risk. The findings could lead to the development of a clinical tool that identifies high-risk adolescents at initial assessment, which would enable improved personalization of treatment and monitoring early on. Ultimately, this work could reduce relapse rates and improve long-term outcomes in the ED field.
Programme
Master's Award: Canada Graduate Scholarships
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
CIHR Funding Decisions Database
Tel que publié
Id
530117
Name
Ryall Stephanie
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Prcname2
Special Cases - Awards Programs
Projecttitle
Using Machine Learning and Precision Medicine to Identify Youth at Risk of Relapsing Eating Disorders
Programtype2
Studentship Programs
Orgnameinp2
University of Ottawa
Cihramount2
$27,000
Cihrequipment2
$0
Programname2
Master's Award: Canada Graduate Scholarships
Keyworddelim
Eating Disorders; Machine Learning; Precision Medicine
Abstract
Background:Eating disorders (ED) are serious mental health conditions with high mortality and relapse rates, often emerging during adolescence. Early intervention with personalized treatment could improve outcomes, but the inability to accurately predict disease course and relapse risk hinders this potential. Application of machine learning (ML) and precision medicine (PM) biological investigations has been proposed as promising strategies to overcome these limitations and enhance ED relapse prediction.Objectives:1.
Assess if ML improves prediction of relapsing ED in adolescents using intake assessment data.2.
Evaluate the added value of PM investigations in predicting relapsing ED.3.
Identify key factors contributing to relapse risk across biopsychosocial domains.Methods:Longitudinal data (2018-2024) from 383 patients in the CHEO ED program, including clinical, biological, psychological, and social metrics collected at intake across all treatment levels will be analyzed. A subset of patients (n=95) will also have PM data including microbiome, inflammatory, endocrine, sleep, smell, and taste testing. Predictive modeling of relapsing ED status will be compared between ML methods (Random Forest and Elastic Net) and logistic regression. Models will be built with and without PM biological data to assess the utility of multi-axial data.Significance:This will be the first study to apply both ML and PM to enhance prediction and understanding of ED relapse risk. The findings could lead to the development of a clinical tool that identifies high-risk adolescents at initial assessment, which would enable improved personalization of treatment and monitoring early on. Ultimately, this work could reduce relapse rates and improve long-term outcomes in the ED field.
Competitiondate
202412
Approvedterm2
1 yr 0 mth
Pinamesdelim
Ryall, Stephanie
Approvedterm
1 an 0 mois
Cihrcontribution2
~ $27,000~ $0~