There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | Vernon Chinchilli |
| Institution: | Penn State College of Medicine |
| Department: | Public Health Sciences |
| Country: | |
| Proposed Analysis: | AS SCIENTIFIC DATA PROBLEMS GROW IN TERMS OF BOTH EXPANDING PARAMETER DIMENSIONS AND SAMPLE SIZES, DIMENSION REDUCTION AND INTEGRATIVE ANALYSIS BECOME CRITICAL CONCEPTS IN PRACTICAL DATA ANALYSIS AND INFERENCE. RECENT ADVANCEMENTS IN HIGH-THROUGHPUT, BIOMEDICAL TECHNOLOGIES HAVE ENABLED THE MEASUREMENT OF MULTIPLE HIGH-DIMENSIONAL OMICS DATA TYPES IN A SINGLE STUDY. FOR EXAMPLE, SUCH STUDIES MIGHT INCLUDE GENOMICS, EPIGENOMICS, TRANSCRIPTOMICS AND METABOLOMICS. EACH OF THESE DATA TYPES PROVIDES A DIFFERENT SNAPSHOT OF THE UNDERLYING BIOLOGICAL SYSTEM, AND COMBINING MULTIPLE DATA TYPES HAS BEEN SHOWN TO BE VERY VALUABLE IN INVESTIGATING DISEASES. INDIVIDUAL COMPONENTS IN THESE DATA ARE FUNCTIONALLY STRUCTURED IN NETWORKS OR PATHWAYS AND INCORPORATION OF SUCH STRUCTURAL INFORMATION CAN IMPROVE ANALYSIS AND LEAD TO BIOLOGICALLY MORE MEANINGFUL RESULTS. MY STUDENT AND I ARE WORKING ON THE FOLLOWING STATISTICAL TOPICS IN THIS REGARD: 1) A GROUP CANONICAL CORRELATION ANALYSIS (GCCA) MODEL; 2) A DEEP LEARNING MODEL OF DYNAMIC GROUP FACTOR ANALYSIS MODEL (DGFA). THE FIRST MODEL (GCCA) CAN EXTRACT THE HIDDEN STRUCTURE AMONG DIFFERENT DATASETS FOR ONE SINGLE STUDY, FOR EXAMPLE, DATA ORIGINATING FROM GENOMICS, IMAGING, EPIGENOMICS, METABOLOMICS, CLINICAL OUTCOMES, ETC. OF PATIENTS (AT LEAST TWO DIFFERENT DATASETS). THE GCCA MODEL CAN EXTRACT THE DEPENDENCIES AMONG THOSE DATASETS. FOR EXAMPLE, IT CAN INFORM THE RESEARCHER AS TO WHICH SUBSETS OF GENES ARE ASSOCIATED WITH SPECIFIC IMAGING FEATURES (IF WE JUST HAVE IMAGING AND GENE EXPRESSION DATA). THE SECOND MODEL ON DEEP LEARNING IS AN EXTENSION OF THE FIRST MODEL TO A LONGITUDINAL STUDY. THE ADVANTAGE OF A LONGITUDINAL DESIGN IS THAT THE RESEARCHER CAN ATTEMPT TO UNDERSTAND THE DISEASE PROPAGATION. WE PROPOSE USING A DEEP LEARNING MODEL BECAUSE WE BELIEVE THAT THE COMPLEX DATASET CAN BE BETTER EXPLAINED USING A NONLINEAR DEEP MODEL AND THE LONGITUDINAL STUDY WILL BE MUCH BETTER TO STUDY THE DISEASE COMPARED TO ONE SINGLE TIME POINT STUDY. |
| Additional Investigators | |
| Investigator's Name: | Lin Qiu |
| Proposed Analysis: | LIN QIU IS MY BIOSTATISTICS PHD STUDENT WHO WILL ASSIST ME IN THE ANALYSES I PROPOSED. |
