| Project/Area Number |
26540112
|
|---|
| Research Category |
Grant-in-Aid for Challenging Exploratory Research
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Tohoku University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Renkei-kenkyūsha) |
ISHIBASHI Tadashi 東北大学, 大学院医学系研究科, 教授 (40151401)
KAWASUMI Yusuke 東北大学, 大学院医学系研究科, 講師 (00513540)
YOSHIZAWA Makoto 東北大学, サイバーサイエンスセンター, 教授 (60166931)
|
|---|
| Research Collaborator |
GUPTA Madan University of Saskatchewan, College of Engineering, Professor Emeritus
HOU Zeng-Guang Chinese Academy of Sciences, Institute of Automation, Professor
BUKOVSKY Ivo Czech Technical University in Prague, Faculty of Mechanical Engineering, Associate Professor
ZHANG Xiaoyong 東北大学, 大学院工学研究科, 助教 (90722752)
|
|---|
| Project Period (FY) |
2014-04-01 – 2017-03-31
|
| Project Status |
Completed (Fiscal Year 2016)
|
| Budget Amount *help |
¥3,640,000 (Direct Cost: ¥2,800,000、Indirect Cost: ¥840,000)
Fiscal Year 2016: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2015: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2014: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
|
|---|
| Keywords | 計算機支援診断システム / 乳房X線撮影 / 機械学習 / 知能情報処理 / 乳房Ｘ線撮影 / 画像認識 |
|---|
| Outline of Final Research Achievements |
In breast cancer screening using mammography, due to the increase of the number of examinees, reading a lot of mammograms became burden for doctors, and it might lead to false detection and unnecessary biopsies. For reducing the work burden of doctors and improving their diagnostic accuracy, computer-aided diagnosis (CAD) systems have been developed. However, it is difficult to design the quantitative features that sufficiently represent the characteristics of abnormalities in mammograms for accurate diagnosis. To solve this problem, we have developed a new CAD system based on a deep learning technique that can extract such features through learning massive data sets. The experimental results showed that diagnostic sensitivity of a typical abnormality was about 90 % and false positive was 20 %. The results demonstrated that the proposed deep learning technique has a potential to be a key strategy for mammographic CAD systems.
|
