Grant: Discovery Grants Program - Individual
University of Guelph
$50,000.00
Development of Artificial Intelligence-based algorithms for Computer-Aided Diagnosis and Prognosis of Cancer
FY 2023-24
Summary
Medical imaging plays a critical role in the diagnosis and prognosis of cancers, such as lung, kidney, and prostate cancer. However, while imaging analysis has advanced in recent years, most medical diagnosis and treatment still heavily relies on manual interpretations of the medical images, which is tedious and subject to observer variability. Recent advancements in artificial intelligence (AI) have created a wealth of novel research opportunities in developing new and fully automated methods for detecting and segmenting anomalies (e.g., cancer) in medical images. AI-based methods have been widely utilized in consumer technology, social media and finance, etc; however, their adaptation to medical image interpretation has been slow. Despite recent intriguing success, AI-based methods still face challenges that impede their efficacy in medical imaging applications, including algorithms that are not well generalized to unseen data, limited image datasets and annotations due to patient privacy concerns, and lack of densely annotations of organs and abnormalities. The long-term goal of this research program is to solve engineering problems in medical image interpretation by developing novel fully automated image analysis algorithms based on machine learning for classification and prediction. The short-term objectives of this program are to: 1) investigate and develop semi-supervised and weakly supervised machine learning methods to utilize unlabelled data towards improving both the accuracy and generalization of computer-aided detection and classification of cancer; 2) investigate and develop privacy preserving decentralized machine learning methods for detection of cancer in medical images; and 3) develop AI-based predictive models for 5-year patient survival using imaging data and heterogeneous data, such as patient characteristics. This research program will develop novel AI-based algorithms that can address key challenge of medical data scarcity in the field, which will allow us to go beyond the state-of-the-art techniques that are usually fully supervised, require large annotated data sets and are, therefore, limited in their utility due to lack of generalization. The proposed semi-supervised, weakly supervised, decentralized learning methods will allow us to harness the power of unused (due to privacy concerns), unlabelled and weakly labelled data towards building more robust, accurate and generalized algorithms for cancer diagnosis and prognosis. Trainees will gain in depth knowledge and expertise in cutting-edge artificial intelligence technologies and become valuable assets to the AI industry in Canada. Furthermore, technological advances in algorithm development may lead to new patentable methods, or even ready-to-use diagnostic software, which can be the basis for a start-up in Canada's growing "MedTech" sector, positioning Canada to be a global leader in the development of automated tools for cancer diagnosis.
Program
Discovery Grants Program - Individual
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
RGPIN-2023-04976
Name nom
Ukwatta, Eranga
Department département
Engineering, School of - Engineering, School of
Organization id
24
Institution établissement
University of Guelph
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
50000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1510
Committee name en
Electrical and Computer Engineering
Committee name fr
Génie électrique et informatique
Area of application code
1102
Area of application group en
Health, education and social services
Area of application group fr
Santé publique, éducation et services sociaux
Area of application en
Biomedical engineering
Area of application fr
Génie biomédical
Research subject code
2514
Research subject group en
Electrical and electronic engineering
Research subject group fr
Digital signal processing
Research subject en
Génie électrique et électronique
Research subject fr
Traitement des signaux numériques
Application title
Development of Artificial Intelligence-based algorithms for Computer-Aided Diagnosis and Prognosis of Cancer
Keyword
Abnormality Detection; Artificial Intelligence; Computer-aided Diagnosis; Federated Learning; Image Processing; Image Segmentation; Machine Learning; Medical Image Analysis; Medical Imaging; Semi-supervised Learning
Application summary
Medical imaging plays a critical role in the diagnosis and prognosis of cancers, such as lung, kidney, and prostate cancer. However, while imaging analysis has advanced in recent years, most medical diagnosis and treatment still heavily relies on manual interpretations of the medical images, which is tedious and subject to observer variability. Recent advancements in artificial intelligence (AI) have created a wealth of novel research opportunities in developing new and fully automated methods for detecting and segmenting anomalies (e.g., cancer) in medical images. AI-based methods have been widely utilized in consumer technology, social media and finance, etc; however, their adaptation to medical image interpretation has been slow. Despite recent intriguing success, AI-based methods still face challenges that impede their efficacy in medical imaging applications, including algorithms that are not well generalized to unseen data, limited image datasets and annotations due to patient privacy concerns, and lack of densely annotations of organs and abnormalities. The long-term goal of this research program is to solve engineering problems in medical image interpretation by developing novel fully automated image analysis algorithms based on machine learning for classification and prediction. The short-term objectives of this program are to: 1) investigate and develop semi-supervised and weakly supervised machine learning methods to utilize unlabelled data towards improving both the accuracy and generalization of computer-aided detection and classification of cancer; 2) investigate and develop privacy preserving decentralized machine learning methods for detection of cancer in medical images; and 3) develop AI-based predictive models for 5-year patient survival using imaging data and heterogeneous data, such as patient characteristics. This research program will develop novel AI-based algorithms that can address key challenge of medical data scarcity in the field, which will allow us to go beyond the state-of-the-art techniques that are usually fully supervised, require large annotated data sets and are, therefore, limited in their utility due to lack of generalization. The proposed semi-supervised, weakly supervised, decentralized learning methods will allow us to harness the power of unused (due to privacy concerns), unlabelled and weakly labelled data towards building more robust, accurate and generalized algorithms for cancer diagnosis and prognosis. Trainees will gain in depth knowledge and expertise in cutting-edge artificial intelligence technologies and become valuable assets to the AI industry in Canada. Furthermore, technological advances in algorithm development may lead to new patentable methods, or even ready-to-use diagnostic software, which can be the basis for a start-up in Canada's growing "MedTech" sector, positioning Canada to be a global leader in the development of automated tools for cancer diagnosis.