| Project/Area Number |
22K15658
|
|---|
| Research Category |
Grant-in-Aid for Early-Career Scientists
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Basic Section 51030:Pathophysiologic neuroscience-related
|
|---|
| Research Institution | Institute of Physical and Chemical Research |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2022-04-01 – 2026-03-31
|
| Project Status |
Granted (Fiscal Year 2023)
|
| Budget Amount *help |
¥4,420,000 (Direct Cost: ¥3,400,000、Indirect Cost: ¥1,020,000)
Fiscal Year 2025: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2024: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2023: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2022: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
|
|---|
| Keywords | segmentation / alzheimer's disease / glial cells / plaques / deep learning / amyloid / image processing / microscopy / 画像情報処理・画像認識 / 分子・細胞・神経生物学 / 認知症疾患 / グリア細胞 / コンピュータビジョン |
|---|
| Outline of Research at the Start |
Our goal is to create an automated tool to conduct image analyses common in Alzheimer's disease research, specifically: quantification of pathological Aβ plaques and glial cells. Automating analyses will reduce human error and bias, thereby improving the reproducibility of Alzheimer's disease research. The tool will be developed in collaboration with neuroscientists to ensure that it is easily understood and useful.
|
| Outline of Annual Research Achievements |
The purpose of the project is to develop an interpretable image processing pipeline that uses deep learning tools to automate image analyses of AD microscopy images. In FY2022 and FY2023, the Research Plan was to focus on image preprocessing and processing of Ab plaques in images.
In the previous year, it became evident that it was necessary to develop methods that can handle images of varying contrast, brightness, hue, etc. To this end, we developed a deep learning meta-network that learns to combine different segmentation maps to generate one that most closely resembles the ground truth label. This work will be published as a short paper in the Medical Imaging with Deep Learning conference this year. The meta-network will be helpful in segmenting AD microscopy datasets, which generally have low signal-to-noise ratio, particularly at advanced stages of disease.
|
| Current Status of Research Progress |
Current Status of Research Progress
3: Progress in research has been slightly delayed.
Reason
Progress is being made in developing deep learning networks for segmentation, as well as automated processing pipelines and data organizing structures. These results have been published in international and local conferences. However, there is some delay in generating specific results for the datasets that we have, and in obtaining more diverse datasets. We aim to apply the tools that we have developed in the previous 2 years in the current fiscal year.
|
| Strategy for Future Research Activity |
For the current fiscal year, we plan to apply the tools that we developed to the AD datasets and present these findings at the Japanese Neuroscience Conference. At the conference, we also plan to recruit imaging datasets from other researchers in the AD field, for the purpose of further developing our pipeline and to encourage collaboration.
|
