Grant: Discovery Launch Supplement
University of Toronto
$12,500.00
Tools and algorithms for accurate longitudinal dose assessment in multi-modal cancer radiation therapy
EF 2024-25
Résumé
Annually, over 30,000 women in North America undergo intensive curative pelvis radiation therapy for various cancers including cervical, endometrial, anal, and vaginal. Moreover, in recent years, the medical community has observed a growing trend towards the use of re-irradiation for recurrent cancer treatments. Definitive RT is the cornerstone for achieving cure in gynecological cancers and is typically administered via two modalities: external beam radiation therapy (EBRT) which utilizes high-energy x-ray and electron beams, external to the patient, and brachytherapy (BT), wherein the radiation source, such as 192-Iridium, is directly placed into the tumor using an applicator. As a result, accurately accumulating radiation doses from these various treatment modalities, whether delivered sequentially over a single course of multi-fractionated radiotherapy or in the context of re-irradiation, has become increasingly crucial. However, the prevailing solutions for dose accumulation, typically done via deformable image registration (DIR), suffer for large uncertainties. These uncertainties arise from significant changes in soft tissue due to the presence or absence of BT applicator with different treatment modalities, physiological changes in bladder and rectum fillings, or biological response of tumor (volume reduction) to treatment. These limitations restrict accurate longitudinal dose accumulation over the course of treatment that can include various radiotherapy modalities or when considering re-irradiation. This, in turn, limits the dose evaluation to the normal healthy tissue and organs.
The integration of deep learning techniques into deformable image registration processes can transform radiation therapy by improving the accuracy and efficiency of aligning multi-modal images and of dose summation across various treatment modalities. Deep learning, a machine learning branch, uses artificial neural networks for automated data feature extraction, making it ideal for complex tasks such as deformable image registration.
The overarching objective of this grant is thus to establish a comprehensive framework for accurate deformable image registration and mapping of radiation doses across different treatment modalities such as EBRT and BT. This will be achieved by: (1) Developing a deep learning-based deformable image registration framework; (2) Developing a realistic female pelvis deformable phantom; (3) Establishing a dosimeter system for experimental validation of phantom dose.
Programme
Discovery Launch Supplement
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
DGECR-2024-00261
Name nom
Serban, Monica
Department département
Radiation Oncology
Organization id
31
Institution établissement
University of Toronto
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
12500
Program id
DGECR
Program name en
Discovery Launch Supplement
Program name fr
Supplément Tremplin vers la découverte
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
3114
Research subject group en
Physics
Research subject group fr
Physique
Research subject en
Biological and Medical Physics
Research subject fr
Physique biologique et médicale
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
Tools and algorithms for accurate longitudinal dose assessment in multi-modal cancer radiation therapy
Keywords
brachytherapy; deformable image registration; dose summation; dose-surface maps; dose-surface histograms; deformable phantom; algorithms; external beam radiotherapy; deep learning methods
Application summary
Annually, over 30,000 women in North America undergo intensive curative pelvis radiation therapy for various cancers including cervical, endometrial, anal, and vaginal. Moreover, in recent years, the medical community has observed a growing trend towards the use of re-irradiation for recurrent cancer treatments. Definitive RT is the cornerstone for achieving cure in gynecological cancers and is typically administered via two modalities: external beam radiation therapy (EBRT) which utilizes high-energy x-ray and electron beams, external to the patient, and brachytherapy (BT), wherein the radiation source, such as 192-Iridium, is directly placed into the tumor using an applicator. As a result, accurately accumulating radiation doses from these various treatment modalities, whether delivered sequentially over a single course of multi-fractionated radiotherapy or in the context of re-irradiation, has become increasingly crucial. However, the prevailing solutions for dose accumulation, typically done via deformable image registration (DIR), suffer for large uncertainties. These uncertainties arise from significant changes in soft tissue due to the presence or absence of BT applicator with different treatment modalities, physiological changes in bladder and rectum fillings, or biological response of tumor (volume reduction) to treatment. These limitations restrict accurate longitudinal dose accumulation over the course of treatment that can include various radiotherapy modalities or when considering re-irradiation. This, in turn, limits the dose evaluation to the normal healthy tissue and organs.
The integration of deep learning techniques into deformable image registration processes can transform radiation therapy by improving the accuracy and efficiency of aligning multi-modal images and of dose summation across various treatment modalities. Deep learning, a machine learning branch, uses artificial neural networks for automated data feature extraction, making it ideal for complex tasks such as deformable image registration.
The overarching objective of this grant is thus to establish a comprehensive framework for accurate deformable image registration and mapping of radiation doses across different treatment modalities such as EBRT and BT. This will be achieved by: (1) Developing a deep learning-based deformable image registration framework; (2) Developing a realistic female pelvis deformable phantom; (3) Establishing a dosimeter system for experimental validation of phantom dose.