2002 Fiscal Year Final Research Report Summary
A basic research on multi-CAD (computer-aided diagnosis) system for 3-dimensional images
| Project/Area Number |
13670926
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Radiation science
|
|---|
| Research Institution | Gifu University |
|---|
Principal Investigator |
FUJITA Hiroshi Gifu University, Graduate School of Medicine, Professor, 大学院・医学研究科, 教授 (10124033)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
HARA Takeshi Graduate Shool of Medicine, Associate Professor, 大学院・医学研究科, 助教授 (10283285)
GOTO Hiroo Faculty of Medicine, Associate Professor, 医学部, 助教授 (80144012)
HOSHI Hiroaki Faculty of Medicine, Professor, 医学部, 教授 (60128395)
KATADA Kazuhiro Fujita Health University, Faculty of Medicine, Professor, 医学部, 教授 (00101684)
ISHIGAKI Takeo Nagoya University, Graduate School of Medicine, Professor, 大学院・医学研究科, 教授 (60094356)
|
|---|
| Project Period (FY) |
2001 – 2002
|
| Keywords | Computer-aided diagnosis / Image analysis / Image Procession / 3-dimentional image processing / Pattern recognition / CAD |
|---|
| Research Abstract |
In this study, the basic researches on developing computer-aided diagnosis (CAD) systems, multi-CADs, for several diagnostic regions and several imaging modalities shown below have been done for two years. These areas of multi-CADs are 1) chest x-ray CT images, 2) abdominal x-ray CT images, 3) brain MR images, 4) abdominal MR images, and 5) breast ultrasound images. In the developments of CAD systems, one of the important issues are the collection of image data with good quality and good quantity and we were able to construct several databases including 50-200 cases on these areas under the collaborative work with many researchers. In the project 1), the scheme for detecting the nodules in the lung area automatically was developed and the algorithms for extracting and recognizing the normal anatomical structures such as bone, bronchus and hilus pulmonis, from torso CT images were written. Preliminary study for classifying the lung field was also performed. In 2), the automated detection of colonic polyps and recognition of liver region were investigated. The quantitative analysis of vertebral body for osteoporosis diagnosis was also studied. In 3), lacunar infarct regions were automatically determined. In 4), liver areas were extracted and automatic classification scheme of liver diseases was developed. In 5), classification algorithms of breast tumors on malignancy was investigated based on 3-dimentional analysis of ultrasound images. In all of these studies new techniques and schemes were successfully developed and the performance was investigated using collected databases. Also, the clinical usefulness of the systems was also discussed. We have successfully developed the effective and useful multi-CAD systems; however there are issues to be solved such as technical problems on the system and necessity of increasing the databases. This suggests that a continuing research is necessary for this multi-CAD project.
|
