2009 Fiscal Year Final Research Report
Development of accurate computer-aided diagnosis of pulmonary nodules in x-ray CT images by using multi-scale morphological recognition
| Project/Area Number |
19500413
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Medical systems
|
|---|
| Research Institution | Tohoku University |
|---|
Principal Investigator |
HOMMA Noriyasu Tohoku University, サイバーサイエンスセンター, 准教授 (30282023)
|
|---|
| Co-Investigator(Renkei-kenkyūsha) |
ISHIBASHI Tadashi 東北大学, 大学院・医学系研究科, 教授 (40151401)
|
|---|
| Research Collaborator |
GUPTA Madan Department of Engineering, University of Saskatchewan, Professor Emeritus
|
|---|
| Project Period (FY) |
2007 – 2009
|
| Keywords | 検査・診断システム / 知的画像診断 |
|---|
| Research Abstract |
In this research project, we developed a new intelligent method for computer aided detection of pulmonary nodules in x-ray CT images. An essential core of the method is to use some pieces of radiologist's professional knowledge to achieve clinically demanded high detection accuracy. The results demonstrated that the method can improve 30% of true positive rate for hardly-detectable nodules and reduce 40% of false positive rate compared to previous ones.
|
