Grant: Idea to Innovation
University of Toronto
$111,625.00
Developing a fully automatic segmentation algorithm for breast MRI
EF 2012-13
Résumé
Magnetic resonance imaging of the breast is a more sensitive method of detecting breast cancer than conventional x-ray mammography and it is now recommended for screening women who are known to be at higher risk of breast cancer. The number of breast MRI exams carried out in Canada and around the world is increasing rapidly and this has lead to a major problem in finding enough trained radiologists to interpret the exams. Each MRI exam contains over 1000 images and reviewing all of the information is time consuming. This has lead to a huge demand for workstations that are able to support the radiologist and streamline the work flow by preparing the images for 3D viewing and using image processing techniques to identify suspicious regions.
We have invented a novel automatic method to isolate the the breast tissue from the unwanted parts of the image. This is important for several reasons. Firstly it allows 3D images to be generated without the clutter of unimportant structures such as the heart obscuring the breast tissue; secondly it allows the volume of the breast to be measured which is essential for calculating breast density, and finally it improves the reliability of computer aided detection and diagnosis algorithms as it removes unimportant regions and provides spatial information.
Although many groups are currently working on this problem our algorithm has some unique advantages. Most significantly it does not rely on the images being acquired with a particular MRI protocol - this is important as protocols vary significantly between North America and Europe and existing algorithms are too specific. Our method is also very fast which means that it can be applied in a clinical context.
The aim of this project is to test the algorithm on a variety of imaging protocols to demonstrate its adaptability, and to develop robust and fast software to carry out the segmentation.
Programme
Idea to Innovation
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
434897-2012
Name nom
Martel, Anne
Department département
Medical Biophysics
Organization id
31
Institution établissement
University of Toronto
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2012
Competition year année de concours
2012
Award amount
111625
Program id
I2IPJ
Program name en
Idea to Innovation
Program name fr
De l'idée à l'innovation
Group en
Research partnerships programs
Group fr
Programmes de partenariat
Committee code
96
Committee name en
RPP Internal Decision Cttee
Committee name fr
Cté de décision interne - PPR
Area of application code
801
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Computer software
Area of application fr
Logiciel
Research subject code
2708
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Image processing
Research subject fr
Traitement des images
Partie
1
Nb partie
1
Application title
Developing a fully automatic segmentation algorithm for breast MRI
Keyword
NSERC
Application summary
Magnetic resonance imaging of the breast is a more sensitive method of detecting breast cancer than conventional x-ray mammography and it is now recommended for screening women who are known to be at higher risk of breast cancer. The number of breast MRI exams carried out in Canada and around the world is increasing rapidly and this has lead to a major problem in finding enough trained radiologists to interpret the exams. Each MRI exam contains over 1000 images and reviewing all of the information is time consuming. This has lead to a huge demand for workstations that are able to support the radiologist and streamline the work flow by preparing the images for 3D viewing and using image processing techniques to identify suspicious regions.
We have invented a novel automatic method to isolate the the breast tissue from the unwanted parts of the image. This is important for several reasons. Firstly it allows 3D images to be generated without the clutter of unimportant structures such as the heart obscuring the breast tissue; secondly it allows the volume of the breast to be measured which is essential for calculating breast density, and finally it improves the reliability of computer aided detection and diagnosis algorithms as it removes unimportant regions and provides spatial information.
Although many groups are currently working on this problem our algorithm has some unique advantages. Most significantly it does not rely on the images being acquired with a particular MRI protocol - this is important as protocols vary significantly between North America and Europe and existing algorithms are too specific. Our method is also very fast which means that it can be applied in a clinical context.
The aim of this project is to test the algorithm on a variety of imaging protocols to demonstrate its adaptability, and to develop robust and fast software to carry out the segmentation.