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: | Jean-Baptiste SCHIRATTI |
| Institution: | OWKIN |
| Department: | OWKIN LAB |
| Country: | |
| Proposed Analysis: | OWKIN is a French company whose goal is to augment medical research by creating novel AI algorithms. Its research lab, OWKIN Lab (led by Gilles Wainrib and Thomas Clozel), focuses on developing new machine learning methods to predict treatment outcomes or disease progression. It benefits from a strong track record of scientific publications, with a recent paper in Nature Medicine (https://www.nature.com/articles/s41591-019-0583-3). The datasets collected throughout clinical studies is often longitudinal and multimodal as they are aimed at describing and characterizing the temporal progression of a disease. The modeling of disease progression from such datasets has been a blooming topic during the last decade. For instance, in the 2019 TADPOLE challenge (https://tadpole.grand-challenge.org/) participants were asked to model the temporal progression of Alzheimer's disease (AD) from multimodal longitudinal datasets. However, many disease progression models still rely on tabular data (laboratory results, feature extracted/computed from images,...). All the models from the prospective TADPOLE challenge are trained on features extracted from T1 brain MRI images and clinical data. Nonlinear registration methods such as LDDMM can be used to bridge this gap and develop disease progression models from shapes or images but they remain computationally very intensive (making them unsuitable for large datasets) and fail to accommodate for multimodal data. To bridge this gap, OWKIN Lab is developing AI algorithms to estimate disease progression from images, combined with other modalities (such as genomics and laboratory/demographics data). In particular, we focus our research on models for structured data (such as variants of Conditional Random Fields and Recurrent Neural Networks). The ADNI data (and TADPOLE data, if still available) shall be used to evaluate the proposed method. No data shall be used for commercial use or in a commercial product. |
| Additional Investigators |
