| Project/Area Number |
16H05913
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (A)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Medical systems
|
|---|
| Research Institution | Oshima National College of Maritime Technology |
|---|
Principal Investigator |
Tachibana Rie 大島商船高等専門学校, その他部局等, 准教授 (90435462)
|
|---|
| Research Collaborator |
YOSHIDA Hiroyuki Massachusetts General Hospital, 3D Imaging Research, Director
NÄPPI Janne J. Massachusetts General Hospital, 3D Imaging Research
|
|---|
| Project Period (FY) |
2016-04-01 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥17,420,000 (Direct Cost: ¥13,400,000、Indirect Cost: ¥4,020,000)
Fiscal Year 2018: ¥3,640,000 (Direct Cost: ¥2,800,000、Indirect Cost: ¥840,000)
Fiscal Year 2017: ¥5,070,000 (Direct Cost: ¥3,900,000、Indirect Cost: ¥1,170,000)
Fiscal Year 2016: ¥8,710,000 (Direct Cost: ¥6,700,000、Indirect Cost: ¥2,010,000)
|
|---|
| Keywords | 深層学習 / 医用システム / CTコロノグラフィ / Dual energy CT / Deep learning / 機械学習 |
|---|
| Outline of Final Research Achievements |
This goal of this project was to develop an electronic colon cleansing system using multiple deep convolutional neural networks (DCNNs) and material decomposition information. Preliminary results suggested that the EC method using multiple DCNNs and material decomposition information can remove residual fecal material from CTC images without generating major EC artifacts in the accuracy of EC over using previous method. However, the EC method using multiple DCNNs and material decomposition information was computationally expensive because it had been designed to analyze 54 cut-plane images at each voxel.
|
| Academic Significance and Societal Importance of the Research Achievements |
提案した医用画像における領域分割のための多列化深層学習は，複数の画像情報を用いて高精度な領域分割を行うことを可能とし，PET-CT画像などの複数モダリティ画像を用いたシステム構築にも応用が可能であると推測する．提案手法により高精度な大腸電子洗浄法が可能となったが，実用化に至るレベルには達しなかった．しかし，本研究による成果は，前処置なしのCT撮像のみによる大腸がん検診の可能性を示唆しており，多くの人が抵抗感なく大腸がん検診を受診できるシステムの実現につながると考える．
|
