2020 Fiscal Year Final Research Report
Polyp 3D Shape Information Acquisition from Endoscopic Images for Computer-Aided Diagnosis
| Project/Area Number |
19K24370
|
|---|
| Research Category |
Grant-in-Aid for Research Activity Start-up
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
1002:Human informatics, applied informatics and related fields
|
|---|
| Research Institution | Chubu University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2019-08-30 – 2021-03-31
|
| Keywords | 内視鏡支援システム / ３次元形状情報獲得 / 医用画像処理 / Deep Learning / 血管領域抽出 / 敵対的生成ネットワーク / Shape from Shading |
|---|
| Outline of Final Research Achievements |
This research established the basis of a new method for acquiring the 3D shape by integrating the latent blood vessel's geometric information and optical constraints.
Construct a neural network for extracting blood vessel regions from endoscopic images and developed an accurate method for analyzing blood vessel structures from the extracted vessel regions. As a result, by focusing on the flow rate of the blood flow that constitutes the blood vessel, the blood vessel's geometric information was obtained correctly by attention to the region consistent with the mainstream and the blood vessel branch. Finally, realized the optical reflection coefficient necessary for the optical shape restoration method was obtained by utilizing the geometric information of the blood vessel as a clue.
Research results have been published in academic journals and international conferences, and a patent application is in progress.
|
| Free Research Field |
人間情報学、応用情報学およびその関連分野
|
| Academic Significance and Societal Importance of the Research Achievements |
悪性腫瘍は，その形態形成に大きく起因し，診断時の重要な情報となる.また，内視鏡による術式は，開腹手術を伴う術式に比べ，身体への負担が軽く，優れるが，大腸は胃や食道に比べて腸管の壁が薄いため技術的な難易度が高く，ポリープの正確な3次元形状情報を獲得することは手術成功率向上に不可分の関係にある.
CT・MRIによる臓器の形状情報獲得手段が挙げられるが，大腸の場合では蠕動運動等の要因から，正確な情報獲得が困難である. 本研究は，一般的な内視鏡画像を対象とし，血管等の潜在的幾何情報と光学的制約を融合させる新たなポリープ3次元形状情報獲得手法の研究開発を行なった点で学術的意義ならびに社会的意義が高い．
|
