Grant: Discovery Grants Program - Individual
Université Laval
$26,000.00
Neuroimaging phenotypes in AD: advanced techniques development
EF 2014-15
Résumé
Background The continuing focus of my NSERC Discovery Grant is the development of methodologies and data/information analysis techniques to support long-term research in aid to diagnostic of neuropsychiatric diseases. The specific applicative domain lies at the intersection of neuroimaging and Alzheimer?s disease and related disorders (ADRD). My first NSERC Discovery Grant was funded in 2009, on the basis that structural neuroimaging could contribute significantly to AD diagnostic/prognostic, provided that certain methodological roadblocks be removed; the latter goal was achieved. This seminal and decisive first award was leveraged 20-fold with 2.55M$ in additional, relevant funding from federal and provincial peer-reviewed programs. Relevant work included 35 peer-reviewed articles/proceedings, four lab reports, three patents applications, and the training/retention of 11 high-quality personnel. Context Recent evidence has demonstrated that a plurality of risk factors, etiological causes and co-morbidities evolve and interact dynamically in ADRD. To further research, there is an urgent need to reduce heterogeneity by unearthing classes of individuals sharing common disease traits and trajectories. We propose to extract neuroimaging phenotypes in an unsupervised manner to uncover commonality in multi-modality data and link the various etiologies to their clinical presentations. However, significant science and engineering (S&E) challenges remain before we can achieve this goal. First, these phenotypes will need to be based on data integrated simultaneously over long timeframes, wide orders of magnitude, and very large samples. A second challenge is the design of machine learning techniques to extract biologically significant features in a coherent and robust fashion from this large assembly of multi-modality neuroimaging data. Finally, a third challenge is the availability of techniques and methods to extract unsupervised phenotypes from the features obtained. Objective The focus of this NSERC Discovery Grant proposal is the continuing development of techniques to remove these obstacles. The aims are: Specific Objective 1: to develop and validate an ADRD ontology for large-scale database integration and processing; Specific Objective 2: to develop and validate multi-acquisition neuromorphological markers, specifically for anatomical and cerebrovascular morphometry; and Specific Objective 3: to develop and validate unsupervised techniques for phenotype identification. We will test our techniques retroactively and prospectively in our medical datawarehouse comprised of multi-modality, multi-timepoint data from 17 different clinical and imaging databases, totaling 4,953 subjects with >26,000 individual MRI datasets and accompanying clinical/cognitive estimates. This is a massive archive, of a magnitude rarely encountered worldwide. Originality / Significance Multi-disciplinary in nature, my research program will provide original developments in computer science, computer vision, biomedical engineering, and mathematics/statistics. It proposes four new opportunities for HQP training, who will gain theoretical and practical expertise in several disciplines in high demand in industry and academia. It will lead to substantial S&E knowledge transfer, including not only manuscripts and presentations at important S&E conferences, but also wider, global dissemination of the ADRD ontology through web services for reuse by other researchers, such as in two large-scale provincial and national Alzheimer?s research projects. In addition to pushing the limits of data/information mining techniques, these new technologies generally enhance the role of neuroimaging as a quantitative biomarker in AD and related disorders.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2014-05804
Name nom
Duchesne, Simon
Department département
Radiology
Organization id
51
Institution établissement
Université Laval
Province en
Québec
Province fr
Québec
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2014
Competition year année de concours
2014
Award amount
26000
Program id
RGPIN
Program name 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
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
1101
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
Human health (including medically-related psychological research)
Area of application fr
Santé publique (y compris recherche psychologique liée à la médecine)
Research subject code
2803
Research subject group en
Artificial intelligence
Research subject group fr
Intelligence artificielle
Research subject en
Knowledge representation
Research subject fr
Répresentation des connaissances
Partie
1
Nb partie
5
Application title
Neuroimaging phenotypes in AD: advanced techniques development
Keyword
NSERC
Application summary
Background The continuing focus of my NSERC Discovery Grant is the development of methodologies and data/information analysis techniques to support long-term research in aid to diagnostic of neuropsychiatric diseases. The specific applicative domain lies at the intersection of neuroimaging and Alzheimer?s disease and related disorders (ADRD). My first NSERC Discovery Grant was funded in 2009, on the basis that structural neuroimaging could contribute significantly to AD diagnostic/prognostic, provided that certain methodological roadblocks be removed; the latter goal was achieved. This seminal and decisive first award was leveraged 20-fold with 2.55M$ in additional, relevant funding from federal and provincial peer-reviewed programs. Relevant work included 35 peer-reviewed articles/proceedings, four lab reports, three patents applications, and the training/retention of 11 high-quality personnel. Context Recent evidence has demonstrated that a plurality of risk factors, etiological causes and co-morbidities evolve and interact dynamically in ADRD. To further research, there is an urgent need to reduce heterogeneity by unearthing classes of individuals sharing common disease traits and trajectories. We propose to extract neuroimaging phenotypes in an unsupervised manner to uncover commonality in multi-modality data and link the various etiologies to their clinical presentations. However, significant science and engineering (S&E) challenges remain before we can achieve this goal. First, these phenotypes will need to be based on data integrated simultaneously over long timeframes, wide orders of magnitude, and very large samples. A second challenge is the design of machine learning techniques to extract biologically significant features in a coherent and robust fashion from this large assembly of multi-modality neuroimaging data. Finally, a third challenge is the availability of techniques and methods to extract unsupervised phenotypes from the features obtained. Objective The focus of this NSERC Discovery Grant proposal is the continuing development of techniques to remove these obstacles. The aims are: Specific Objective 1: to develop and validate an ADRD ontology for large-scale database integration and processing; Specific Objective 2: to develop and validate multi-acquisition neuromorphological markers, specifically for anatomical and cerebrovascular morphometry; and Specific Objective 3: to develop and validate unsupervised techniques for phenotype identification. We will test our techniques retroactively and prospectively in our medical datawarehouse comprised of multi-modality, multi-timepoint data from 17 different clinical and imaging databases, totaling 4,953 subjects with >26,000 individual MRI datasets and accompanying clinical/cognitive estimates. This is a massive archive, of a magnitude rarely encountered worldwide. Originality / Significance Multi-disciplinary in nature, my research program will provide original developments in computer science, computer vision, biomedical engineering, and mathematics/statistics. It proposes four new opportunities for HQP training, who will gain theoretical and practical expertise in several disciplines in high demand in industry and academia. It will lead to substantial S&E knowledge transfer, including not only manuscripts and presentations at important S&E conferences, but also wider, global dissemination of the ADRD ontology through web services for reuse by other researchers, such as in two large-scale provincial and national Alzheimer?s research projects. In addition to pushing the limits of data/information mining techniques, these new technologies generally enhance the role of neuroimaging as a quantitative biomarker in AD and related disorders.