Construction of 3D pathological images and registration to MR image for multimodal analysis of brain tumor

2015 Fiscal Year Final Research Report

• Project/Area Number: 25560189
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Biomedical engineering/Biomaterial science and engineering
• Research Institution: Chiba University
• Principal Investigator: Haneishi Hideaki 千葉大学, フロンティア医工学センター, 教授 (20228521)
• Co-Investigator(Renkei-kenkyūsha): NAKAGUCHI Toshiya 千葉大学, フロンティア医工学センター, 准教授 (20361412)
• Project Period (FY): 2013-04-01 – 2016-03-31
• Keywords: マルチモダリティ / マルチスケール / 病理画像 / MR像 / 画像レジストレーション / テクスチャ解析 / 光学像

Outline of Final Research Achievements

Recently, expectations for multi scale analysis in medical imaging are growing. Prior to such analysis, high quality alignment between images are needed. We propose registration methods. Conventional metrics such as mutual information cannot evaluate similarity between images appropriately when image resolutions are extremely different. In this study, a texture analysis technique was introduced to image registration. Image acquisition experiment using a pig brain was conducted. We confirmed that difference between gray and white matters in pathological image was clearly enhanced and image registration between pathological images and MR image was successfully conducted. We also propose a registration method that mediates optical image of tissue surface. Tissue optical image and pathological image is taken from pig’s brain and alignment between optical image and pathological image by affine transformation is carry out.

Free Research Field

医用画像工学