Grant: Discovery Grants Program - Individual
University of Manitoba
$38,000.00
Inverse Problems in Medical Imaging: Design, Development and Evaluation of novel diagnostic imaging systems and methods.
FY 2024-25
Summary
Medical imaging technology has improved health care but has limitations. Breast Cancer (BC) mortality in marginalised northern communities and low and middle-income countries is higher, in part due to limited access to BC screening. Positron Emission Tomography (PET) is established for diagnostic imaging and radiotherapy planning but needs greater quantitative accuracy, improved attenuation correction and cost-effectiveness at lower doses. This multidisciplinary research program develops and assesses novel imaging applications that improve image quality, patient safety and comfort, and offers better access to care. Our diverse student population brings inherent strengths to the program, while the research provides opportunities for these highly qualified personnel (HQP) in Physics and Engineering to develop in-demand skills in image processing, machine learning (ML), big-data analysis, 3D printing and RF electronics.
As a leader in the experimental use of radar-based microwave techniques to detect BC, we have shown an accuracy approaching that of x-ray mammography using experimental breast phantoms. In PET, our previous work demonstrated that using scattered photons improves image quality for a given dose and enables anatomic details to be reconstructed without CT or MRI. PET improvements need good spatial and energy resolution, and for microwave radar systems, good spatial and temporal linearity, high sensitivity, and a wide frequency bandwidth. We are developing novel analytical and ML reconstructions to minimize the impact of system deficiencies and reduce cost while maximizing image and data quality.
Two main objectives in the current scope of the research program are (1) the development of novel PET image reconstruction algorithms of functional (activity uptake) and anatomical (electron density) images using Generative Adversarial Networks and evaluating these using simulated and experimental PET data and (2) analysis, optimization and improvement of our microwave imaging systems and breast-phantom architectures while advancing machine-learning-based and analytic microwave radar imaging reconstruction algorithms using enhanced physics modelling to improve the sensitivity and specificity of BC detection.
Each objective will include theoretical (T), computational (C) and experimental (E) components, with the HQP working in a collaborative "intrapreneurship" environment with academic and industrial mentors. Experimental data will be obtained from our prototype microwave BC systems, while synthetic data will be generated using Monte Carlo simulations and EM simulation software. Funding from various sources will be used to train at least 6 PhD, 6 MSc, and 20 UGS over five years. The training will blend Physics and Engineering education and skills to develop well-rounded, industry ready HQP in the natural sciences and engineering, inventing, and evaluating disruptive imaging devices using innovative technology to reduce screening inequities.
Program
Discovery Grants Program - Individual
Recipient
Location
Manitoba, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
RGPIN-2023-04725
Name nom
Pistorius, Stephen
Department département
Physics and Astronomy
Organization id
19
Institution établissement
University of Manitoba
Province en
Manitoba
Province fr
Manitoba
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2023
Award amount
38000
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
1505
Committee name en
Physics
Committee name fr
Physique
Area of application code
518
Area of application group en
Manufacturing processes and products
Area of application group fr
Procédés et produits industriels
Area of application en
Medical equipment and apparatus
Area of application fr
Appareils et instruments médicaux
Research subject code
3100
Research subject group en
Physics
Research subject group fr
Physique
Research subject en
Physics
Research subject fr
Physique
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Inverse Problems in Medical Imaging: Design, Development and Evaluation of novel diagnostic imaging systems and methods.
Keywords
diagnostic imaging; microwave radar imaging; positron emission tomography; scatter imaging; image reconstruction; machine learning; quality metrics; breast imaging
Application summary
Medical imaging technology has improved health care but has limitations. Breast Cancer (BC) mortality in marginalised northern communities and low and middle-income countries is higher, in part due to limited access to BC screening. Positron Emission Tomography (PET) is established for diagnostic imaging and radiotherapy planning but needs greater quantitative accuracy, improved attenuation correction and cost-effectiveness at lower doses. This multidisciplinary research program develops and assesses novel imaging applications that improve image quality, patient safety and comfort, and offers better access to care. Our diverse student population brings inherent strengths to the program, while the research provides opportunities for these highly qualified personnel (HQP) in Physics and Engineering to develop in-demand skills in image processing, machine learning (ML), big-data analysis, 3D printing and RF electronics.
As a leader in the experimental use of radar-based microwave techniques to detect BC, we have shown an accuracy approaching that of x-ray mammography using experimental breast phantoms. In PET, our previous work demonstrated that using scattered photons improves image quality for a given dose and enables anatomic details to be reconstructed without CT or MRI. PET improvements need good spatial and energy resolution, and for microwave radar systems, good spatial and temporal linearity, high sensitivity, and a wide frequency bandwidth. We are developing novel analytical and ML reconstructions to minimize the impact of system deficiencies and reduce cost while maximizing image and data quality.
Two main objectives in the current scope of the research program are (1) the development of novel PET image reconstruction algorithms of functional (activity uptake) and anatomical (electron density) images using Generative Adversarial Networks and evaluating these using simulated and experimental PET data and (2) analysis, optimization and improvement of our microwave imaging systems and breast-phantom architectures while advancing machine-learning-based and analytic microwave radar imaging reconstruction algorithms using enhanced physics modelling to improve the sensitivity and specificity of BC detection.
Each objective will include theoretical (T), computational (C) and experimental (E) components, with the HQP working in a collaborative "intrapreneurship" environment with academic and industrial mentors. Experimental data will be obtained from our prototype microwave BC systems, while synthetic data will be generated using Monte Carlo simulations and EM simulation software. Funding from various sources will be used to train at least 6 PhD, 6 MSc, and 20 UGS over five years. The training will blend Physics and Engineering education and skills to develop well-rounded, industry ready HQP in the natural sciences and engineering, inventing, and evaluating disruptive imaging devices using innovative technology to reduce screening inequities.