Unraveling submanifold properties of diffusion MRI signal models and its application
| Reference No. | 2026a033 |
|---|---|
| Type/Category | Grant for General Research-Short-term Joint Research |
| Title of Research Project | Unraveling submanifold properties of diffusion MRI signal models and its application |
| Principal Investigator | Yoshitaka Masutani（Tohoku University Graduate School of Medicine・Professor） |
| Research Period | |
| Keyword(s) of Research Fields | medical image, diffusion MRI, signal model, manifold |
| Abstract for Research Report | Diffusion MRI (dMRI) can obtain information of fine structures and functions in living tissues and is based on measuring water molecule diffusion. By applying a parametric model suited to the purpose to a set of diffusion-weighted images (DWIs) acquired with varying gradient field strengths and directions for diffusion measurement, voxel-level biological information can be obtained as model parameters. For instance, signal models incorporating parameters for general description of physical phenomenon or anatomical geometries are known, such as the diffusion tensor and nerve fiber diameter. Biological information parameters are estimated through fitting based on the imaging settings used and the measured DWI signal values. However, these parameters are often unreliable due to noise present in the DWI, prompting multifaceted research into robust parameter estimation. As a continuation of the previous year's work, this project aims to clarify the properties of the submanifold formed by the parametric signal model of dMRI in the signal value space and to conduct various investigations toward robust parameter estimation. In the previous year, we deepened and shared understanding of the issues among researchers with different backgrounds while also exchanging diverse opinions with researchers in related fields through public seminars. This year, based on the previous year's results, we will begin investigations for developing specific methods. This is expected to contribute not only to parameter estimation but also to improving imaging methods, assessing the reliability of estimated parameters, and enhancing diagnostic accuracy. |
| Organizing Committee Members (Workshop) Participants (Short-term Joint Usage) |
Shizuo KAJI（Kyoto University・Professor） Tomoki UDA（Nanzan University・Associate Professor） Noriyuki HAMADA（Kyushu University・Project Assistant Professor） Hayato ITOH（Fukuoka University・Assistant Professor） |