2021 Fiscal Year Research-status Report
A convolutional neural network based approach for generating full PET/CT image series from shorter scan time
| Project/Area Number |
19K20685
|
|---|
| Research Institution | National Institutes for Quantum Science and Technology |
|---|
Principal Investigator |
BhusalChhatkuli Ritu 国立研究開発法人量子科学技術研究開発機構, 量子医科学研究所 分子イメージング診断治療研究部, 博士研究員 (50836591)
|
|---|
| Project Period (FY) |
2019-04-01 – 2023-03-31
|
| Keywords | Neural Network / PET/CT / dynamic images / RNN / Prediction |
|---|
| Outline of Annual Research Achievements |
-Continuing the acquisition of patient data from the hospital, data preprocessing and mining.
-Trying with several Neural Network methods ( convolutional and recurrent) for creating the prediction model for predicting delayed images from the early and dynamic images.
-Successfully predicting delayed images, its quantitative and qualitative analysis.
-Summarizing the results, its summary and completing the paper.
-Presented a part of research result to JAMIT (Japan society of Medical Imaging Technology)
|
| Current Status of Research Progress |
Current Status of Research Progress
3: Progress in research has been slightly delayed.
Reason
Changes of approach to obtain better result which delayed result summary and paper writing. C
|
| Strategy for Future Research Activity |
-Submitting the completed paper for publication in good journal.
-Further research improvement with the new set of data.
|
| Causes of Carryover |
Covid 19 prevented me to go to several business trips and society meetings, Need to attend several society meetings, pending journal publications and some GPU machines to be bought for further research analysis.
|
Research Products
(1 results)
