Grant: Discovery Grants Program - Individual
Université Laval
$30,000.00
Neuroimaging in dementias: advanced technique development
FY 2011-12
Summary
My long-term research is dedicated to developing groundbreaking analysis techniques for image-based computer-aided diagnosis (CAD) within the main applicative context of dementias (e.g. Alzheimer's (AD)). The focus of this NSERC Discovery research program is to develop and validate robust, automated magnetic resonance imaging (MRI) feature extraction methods, combined with novel high-dimensional data analysis techniques, to increase CAD accuracy. Specific Objective 1: to develop and validate a) image harmonization techniques; b) automated large scale image quality control methods; and c) a control-theory framework for image processing pipeline optimization, for the extraction of robust, automated MRI features in multi-scanner data. The targeted performance is to achieve <5% inter-scanner intensity and registration features variability on the human scan-rescan data, with optimal pipelines and embedded automated quality control resulting in <5% failures. Specific Objective 2: to develop and validate a) nonlinear high-dimensional models; b) quantum classification techniques; and c) an unsupervised learning framework for phenotype identification from large-scale datasets, for the generation of a novel high-dimensional analysis framework. The targeted performance is classification of subjects according to multi-modality phenotypes with higher accuracy than previous approaches. We will base these efforts on our medical data warehouse comprised of more than 2,404 subjects 20 to 96 years of age, translating to more than 6,000 individual MRI datasets and accompanying clinical/cognitive estimates. We know of no other database of this magnitude in the world. With potential for multiple graduate-level projects, this multidisciplinary science and engineering research program strongly benefits the training of HQP. Its successful completion will provide a framework for the analysis, data mining and classification of massive, high-dimensional data, increasing the performance of computer aid to diagnosis in AD.
Program
Discovery Grants Program - Individual
Recipient
Location
Quebec, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
372045-2009
Name nom
Duchesne, Simon
Department département
Radiologie
Organization id
51
Institution établissement
Université Laval
Province en
Québec
Province fr
Québec
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2011
Competition year année de concours
2009
Award amount
30000
Program id
RGPIN
Program nane en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
1079
Committee name en
Electrical and Computer Engineering
Committee name fr
Génie électrique et de l'informatique
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
1900
Research subject group en
Biomedical engineering
Research subject group fr
Génie biomédical
Research subject en
Biomedical engineering
Research subject fr
Génie biomédical
Partie
3
Nb partie
5
Application title
Neuroimaging in dementias: advanced technique development
Keyword
NSERC
Application summary
My long-term research is dedicated to developing groundbreaking analysis techniques for image-based computer-aided diagnosis (CAD) within the main applicative context of dementias (e.g. Alzheimer's (AD)). The focus of this NSERC Discovery research program is to develop and validate robust, automated magnetic resonance imaging (MRI) feature extraction methods, combined with novel high-dimensional data analysis techniques, to increase CAD accuracy. Specific Objective 1: to develop and validate a) image harmonization techniques; b) automated large scale image quality control methods; and c) a control-theory framework for image processing pipeline optimization, for the extraction of robust, automated MRI features in multi-scanner data. The targeted performance is to achieve <5% inter-scanner intensity and registration features variability on the human scan-rescan data, with optimal pipelines and embedded automated quality control resulting in <5% failures. Specific Objective 2: to develop and validate a) nonlinear high-dimensional models; b) quantum classification techniques; and c) an unsupervised learning framework for phenotype identification from large-scale datasets, for the generation of a novel high-dimensional analysis framework. The targeted performance is classification of subjects according to multi-modality phenotypes with higher accuracy than previous approaches. We will base these efforts on our medical data warehouse comprised of more than 2,404 subjects 20 to 96 years of age, translating to more than 6,000 individual MRI datasets and accompanying clinical/cognitive estimates. We know of no other database of this magnitude in the world. With potential for multiple graduate-level projects, this multidisciplinary science and engineering research program strongly benefits the training of HQP. Its successful completion will provide a framework for the analysis, data mining and classification of massive, high-dimensional data, increasing the performance of computer aid to diagnosis in AD.