Tensor Sparse Coding for Temporal and Spatial Feature Extraction and Classification of Liver Lesions in Multi-phase CT Images
Project/Area Number |
18H03267
|
Research Category |
Grant-in-Aid for Scientific Research (B)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Review Section |
Basic Section 61010:Perceptual information processing-related
|
Research Institution | Ritsumeikan University |
Principal Investigator |
CHEN Yen-Wei 立命館大学, 情報理工学部, 教授 (60236841)
|
Co-Investigator(Kenkyū-buntansha) |
岩本 祐太郎 立命館大学, 情報理工学部, 助教 (30779054)
韓 先花 山口大学, 大学院創成科学研究科, 准教授 (60469195)
古川 顕 東京都立大学, 人間健康科学研究科, 教授 (80199421)
金崎 周造 滋賀医科大学, 医学部, 非常勤講師 (90464180)
|
Project Period (FY) |
2018-04-01 – 2021-03-31
|
Project Status |
Completed (Fiscal Year 2022)
|
Budget Amount *help |
¥15,600,000 (Direct Cost: ¥12,000,000、Indirect Cost: ¥3,600,000)
Fiscal Year 2020: ¥3,640,000 (Direct Cost: ¥2,800,000、Indirect Cost: ¥840,000)
Fiscal Year 2019: ¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Fiscal Year 2018: ¥7,800,000 (Direct Cost: ¥6,000,000、Indirect Cost: ¥1,800,000)
|
Keywords | テンソルスパース表現法 / 多時相C T画像 / 肝腫瘍性病変 / 時空間特徴抽出 / 類似症例検索 / パイロット試験 / 肝臓腫瘍病変 / 多重線型理論 / スパース符号化法 / 計算機診断支援 / 深層学習 / 肝臓腫瘍の分類 / 多重線型代数 / スパース表現法 / テンソル / 多時相CT画像 / 肝腫瘍性病変症例検索 / 診断支援 / テンソルスパース表現 / 肝臓のセグメンテーション / 鑑別と分類 / 時空間特徴 |
Outline of Final Research Achievements |
The purpose of this research project is to develop a content based medical image retrieval (CBMIR) for assisting radiologists to detect and characterize focal liver lesions (FLLs) using multi-phase CT images. We proposed a tensor sparse representation method, which treats the multi-phase CT image as a tensor, to effectively extract temporal and spatial features of multi-phase CT images to provide doctors medical cases more relevant to the query one. The diagnosis accuracy is improved to more than 90%. We also developed a CBMIR system for clinical use. Furthermore, we conducted a comparative experiments to evaluate the effectiveness of the proposed method. In addition to conventional computer experiments, we also conducted a pilot trial experiments. Six doctors joined our pilot trial and experimental results show that both diagnosis accuracy and confidence were significantly improved by using our system.
|
Academic Significance and Societal Importance of the Research Achievements |
本研究の学術的意義は以下の4点である;①多重線形代数の枠組でテンソルベーススパース表現法を理論的に開発した;②多時相CT画像を一つのテンソルとして取り扱い、テンソルスパース表現法によって時相間の共起を考慮した時空間特徴抽出法を開発した;③世界初肝腫瘍多時相CT画像データベースを構築した;④提案法の有効性の検証に、従来の計算機実験だけではなく医師によるパイロット試験も実施した。また、本研究提案法の確立により、がんをはじめとする様々な疾患の診断精度が向上し、患者の生存率の向上に寄与できることから、社会的な効果は極めて大きい。
|
Report
(4 results)
Research Products
(64 results)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
[Presentation] Multi-Stream Scale-Insensitive Convolutional and Recurrent Neural Networks for Liver Tumor Detection in Dynamic Ct Images2019
Author(s)
Dong Liang, Ruofeng Tong, Jian Wu, Lanfen Lin, Xiao Chen, Hongjie Hu, Qiaowei Zhang, Qingqing Chen, Yutaro Iwamoto, Xianhua Han, Yen-Wei Chen
Organizer
2019 IEEE International Conference on Image Processing (IEEE ICIP 2019)
Related Report
Int'l Joint Research
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-