Grant: Doctoral Research Award: Canada Graduate Scholarships
University of Toronto
$105,000.00
Improving Prognostication following Traumatic Brain Injury using a Population-Based Deep Learning Approach Incorporating Automated Acute Brain Imaging Analysis
EF 2022-23
Résumé
Traumatic brain injury (TBI) is a major reason for disability, lost quality of life, and mortality with over 69 million cases worldwide each year. TBI affects individuals across the entire spectrum of age; TBI patients may develop other psychiatric or medical problems and often experience shorter lifespans with more dependence on health care. When patients have a head injury, there is a highly variable pattern of brain injury progression. Some severely brain-injured people may progress to independent living while other mildly injured people may develop chronic difficulties typified by medical disability and difficulty finding gainful employment. Predicting which patients will follow which patterns remains challenging for clinicians and patient families involved in TBI care. Brain imaging with computed tomography (CT) has become an important part of assessing TBI patients, however we are still not able to provide accurate outcome prediction following acute brain injury. Our group has developed a method to automatically extract important imaging features using machine learning from brain imaging. We will build upon these models to develop new systems that assess important factors for TBI such as the amount of pressure in the brain, the amount of blood and location of blood. We will then combine data from over 6,000 scans with health outcome data for the population of brain injured patients in Ontario over the last decade. Statistical tests and epidemiological methods will be used to compare the short-term and long-term outcomes. Short-term outcomes include how different TBI imaging patterns affect the need for intensive care unit stay, surgery, early death or disability. Long-term outcomes will also be compared to early post-injury brain imaging; specific outcomes of interest include whether brain-injured patients have increased health resource utilization, return to work after 1 and 5 years, and what proportion become functionally independent at 1 and 5 years.
Programme
Doctoral Research Award: Canada Graduate Scholarships
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
CIHR Funding Decisions Database
Tel que publié
Id
475733
Name
Malhotra Armaan K
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Health Services and Policy Research
Theme2
Health systems / services
Prcname2
Doctoral Research Awards - B
Projecttitle
Improving Prognostication following Traumatic Brain Injury using a Population-Based Deep Learning Approach Incorporating Automated Acute Brain Imaging Analysis
Programtype2
Studentship Programs
Orgnameinp2
University of Toronto
Cihramount2
$105,000
Cihrequipment2
$0
Programname2
Doctoral Research Award: Canada Graduate Scholarships
Keyworddelim
Epidemiology; Health Services Research; Imaging; Machine Learning; Nervous System; Outcomes Research; Prediction; Traumatic Brain Injury
Abstract
Traumatic brain injury (TBI) is a major reason for disability, lost quality of life, and mortality with over 69 million cases worldwide each year. TBI affects individuals across the entire spectrum of age; TBI patients may develop other psychiatric or medical problems and often experience shorter lifespans with more dependence on health care. When patients have a head injury, there is a highly variable pattern of brain injury progression. Some severely brain-injured people may progress to independent living while other mildly injured people may develop chronic difficulties typified by medical disability and difficulty finding gainful employment. Predicting which patients will follow which patterns remains challenging for clinicians and patient families involved in TBI care. Brain imaging with computed tomography (CT) has become an important part of assessing TBI patients, however we are still not able to provide accurate outcome prediction following acute brain injury. Our group has developed a method to automatically extract important imaging features using machine learning from brain imaging. We will build upon these models to develop new systems that assess important factors for TBI such as the amount of pressure in the brain, the amount of blood and location of blood. We will then combine data from over 6,000 scans with health outcome data for the population of brain injured patients in Ontario over the last decade. Statistical tests and epidemiological methods will be used to compare the short-term and long-term outcomes. Short-term outcomes include how different TBI imaging patterns affect the need for intensive care unit stay, surgery, early death or disability. Long-term outcomes will also be compared to early post-injury brain imaging; specific outcomes of interest include whether brain-injured patients have increased health resource utilization, return to work after 1 and 5 years, and what proportion become functionally independent at 1 and 5 years.
Competitiondate
202211
Approvedterm2
3 yrs 0 mth
Pinamesdelim
Malhotra, Armaan K
Approvedterm
3 ans 0 mois
Cihrcontribution2
~ $105,000~ $0~