Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Doctoral Research Award: Canada Graduate Scholarships

University of Toronto

$120,000.00

A Machine Learning Algorithm that Accurately Predicts Spontaneous Recovery in Liver Disease: a clinical tool to optimise organ allocation in liver transplantation

FY 2024-25

Summary
The number of Canadians living with liver disease is increasing. Liver transplantation is often the only life-saving intervention available to people with severe liver disease. However, the demand for liver transplantation far outpaces the supply of donate organs in Canada and across the world. Consequently, 30% of people eligible for liver transplant will die while waiting for an organ. Nonetheless, the liver possesses a capacity to heal itself, and some patients can recover spontaneously from their disease. However, clinicians cannot easily predict which patient can recover on their own and which patient will die without liver transplant. Thus, some patients may end up receiving a liver transplant when spontaneous recovery could have been a possibility, leading to poor utilization of a rare resource. There is currently no statistical model that can accurately predict spontaneous recovery in patients with liver disease, possibly due to the complexity of this clinical question. Machine learning algorithms have proven to be more adept at inferring relationships between complex variables compared to conventional statistical models. Therefore, we hypothesise that a machine learning model can accurately predict spontaneous recovery of liver disease. We propose to build two predictive models as part of a PhD thesis, one specially designed for acute liver failure, and one designed for chronic liver disease. These models will test and validated using the Scientific Registry of Transplant Recipients dataset, a large dataset of 200,000 liver transplant candidates from the United States. Additional validation will be performed on the European Liver Transplant Registry. The goal of this tool is to make organ transplant more efficient and help physicians make more informed decisions regarding the care of their patients.
Program
Doctoral Research Award: Canada Graduate Scholarships
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
CIHR Funding Decisions Database

As published

Id
518300
Name
Zhao Xun
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Health Services and Policy Research
Theme2
Clinical
Prcname2
Doctoral Research Awards - B
Projecttitle
A Machine Learning Algorithm that Accurately Predicts Spontaneous Recovery in Liver Disease: a clinical tool to optimise organ allocation in liver transplantation
Programtype2
Studentship Programs
Orgnameinp2
University of Toronto
Cihramount2
$120,000
Cihrequipment2
$0
Programname2
Doctoral Research Award: Canada Graduate Scholarships
Keyworddelim
Liver Disease; Liver Regeneration; Liver Transplantation; Machine Learning; Organ Allocation
Abstract
The number of Canadians living with liver disease is increasing. Liver transplantation is often the only life-saving intervention available to people with severe liver disease. However, the demand for liver transplantation far outpaces the supply of donate organs in Canada and across the world. Consequently, 30% of people eligible for liver transplant will die while waiting for an organ. Nonetheless, the liver possesses a capacity to heal itself, and some patients can recover spontaneously from their disease. However, clinicians cannot easily predict which patient can recover on their own and which patient will die without liver transplant. Thus, some patients may end up receiving a liver transplant when spontaneous recovery could have been a possibility, leading to poor utilization of a rare resource. There is currently no statistical model that can accurately predict spontaneous recovery in patients with liver disease, possibly due to the complexity of this clinical question. Machine learning algorithms have proven to be more adept at inferring relationships between complex variables compared to conventional statistical models. Therefore, we hypothesise that a machine learning model can accurately predict spontaneous recovery of liver disease. We propose to build two predictive models as part of a PhD thesis, one specially designed for acute liver failure, and one designed for chronic liver disease. These models will test and validated using the Scientific Registry of Transplant Recipients dataset, a large dataset of 200,000 liver transplant candidates from the United States. Additional validation will be performed on the European Liver Transplant Registry. The goal of this tool is to make organ transplant more efficient and help physicians make more informed decisions regarding the care of their patients.
Competitiondate
202411
Approvedterm2
3 yrs 0 mth
Pinamesdelim
Zhao, Xun
Approvedterm
3 ans 0 mois
Cihrcontribution2
~ $120,000~ $0~