Objective Evaluation Method for Optical Colonoscopy Images using Image Recognition Techniques

2014 Fiscal Year Final Research Report

• Project/Area Number: 24700186
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Perception information processing/Intelligent robotics
• Research Institution: National Institute of Advanced Industrial Science and Technology
• Principal Investigator: NOSATO Hirokazu 独立行政法人産業技術総合研究所, 情報技術研究部門, 主任研究員 (40435764)
• Research Collaborator: SUZUKI Yasuo ; AOKI Hiroshi ; TAKEUCHI Ken
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 医用画像認識 / 大腸内視鏡 / 特徴抽出 / 画像処理 / 診断支援

Outline of Final Research Achievements

This research aimed to establish a new objective evaluation method for Ulcerative Colitis (UC) with optical colonoscopy images using image recognition techniques. In the proposed method, the higher-order local auto-correlation method and the multivariate data analysis method were adopted to distinguish between normal and abnormal colonic mucosa. In addition, a color conversion techniques was also adopted to enhance the ability to efficiently observe the colon condition. Using the objective evaluation method, classification method according to UC severity and content-based image retrieval method for colonoscopy images were developed in this research.

Free Research Field

医用画像認識