Canada Spends Logo

Explore our Budget 2026 analysis

Grant: CIHR Fellowship

University of Toronto

$220,000.00

Using machine learning to predict outcomes following major vascular surgery

FY 2022-23

Summary
Background: Vascular surgery patients are at high risk for complications given their significant comorbidities and the nature of their operations involving major vascular systems. Up to 30% of patients undergoing vascular surgery will suffer an adverse event, which remains among the highest for all surgical patients. Current tools for predicting post-operative outcomes are limited by the need to manually input data, non-specificity to high-risk surgical populations, and lack of long-term outcome predictions. Our goal is to develop automated machine learning (ML) algorithms that use pre-operative patient characteristics to predict peri-operative and long-term outcomes following major vascular surgery. We plan to achieve this goal at the local, provincial, and national/international levels. Aims: Aim 1 (local): Develop a ML algorithm using electronic health records data from St. Michael's Hospital (463-bed academic hospital affiliated with the University of Toronto) to predict peri-operative complications following major vascular surgery. Aim 2 (provincial): Develop a ML algorithm using population-level data from the Ontario Institute of Clinical Evaluative Sciences (ICES) linked administrative health databases to predict long-term outcomes following major vascular surgery. Aim 3 (national/international): Develop a ML algorithm using North American registry data from the Society for Vascular Surgery Vascular Quality Initiative Database to predict peri-operative and long-term outcomes following major vascular surgery. Significance: Through this research, we will develop rigorous and versatile ML algorithms applicable to patients locally, provincially, across Canada, and around the world. These risk prediction tools can be deployed into clinical practice to identify high-risk patients and mitigate adverse events by guiding patient selection for surgical interventions and peri-operative management, ultimately improving outcomes for vascular surgery patients.
Program
CIHR Fellowship
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
CIHR Funding Decisions Database

As published

Id
471859
Name
Li Ben
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Circulatory and Respiratory Health
Theme2
Clinical
Prcname2
Health Professional Awards - Fellowships
Projecttitle
Using machine learning to predict outcomes following major vascular surgery
Programtype2
Fellowship Programs
Orgnameinp2
University of Toronto
Cihramount2
$220,000
Cihrequipment2
$0
Programname2
CIHR Fellowship
Keyworddelim
Machine Learning; Outcomes; Risk Prediction; Vascular Surgery
Abstract
Background: Vascular surgery patients are at high risk for complications given their significant comorbidities and the nature of their operations involving major vascular systems. Up to 30% of patients undergoing vascular surgery will suffer an adverse event, which remains among the highest for all surgical patients. Current tools for predicting post-operative outcomes are limited by the need to manually input data, non-specificity to high-risk surgical populations, and lack of long-term outcome predictions. Our goal is to develop automated machine learning (ML) algorithms that use pre-operative patient characteristics to predict peri-operative and long-term outcomes following major vascular surgery. We plan to achieve this goal at the local, provincial, and national/international levels. Aims: Aim 1 (local): Develop a ML algorithm using electronic health records data from St. Michael's Hospital (463-bed academic hospital affiliated with the University of Toronto) to predict peri-operative complications following major vascular surgery. Aim 2 (provincial): Develop a ML algorithm using population-level data from the Ontario Institute of Clinical Evaluative Sciences (ICES) linked administrative health databases to predict long-term outcomes following major vascular surgery. Aim 3 (national/international): Develop a ML algorithm using North American registry data from the Society for Vascular Surgery Vascular Quality Initiative Database to predict peri-operative and long-term outcomes following major vascular surgery. Significance: Through this research, we will develop rigorous and versatile ML algorithms applicable to patients locally, provincially, across Canada, and around the world. These risk prediction tools can be deployed into clinical practice to identify high-risk patients and mitigate adverse events by guiding patient selection for surgical interventions and peri-operative management, ultimately improving outcomes for vascular surgery patients.
Competitiondate
202210
Approvedterm2
4 yrs 0 mth
Pinamesdelim
Li, Ben
Approvedterm
4 ans 0 mois
Cihrcontribution2
~ $220,000~ $0~