内容紹介
人工知能の発展に伴い、画像診断支援機能も進化しつつある。医療画像の場合は、そのまま人工知能に活用できる症例データが不足している。医療画像に対しては一般に、前処理、能動学習、転移学習などの工夫で精度を上げている状況を説明するとともに、弊社で取り組んでいる脳動脈瘤検知アルゴリズムの研究についても紹介する。
Due to advancements in AI-based image processing technology, we are amidst a revolutionary change in its applications in supporting medical image diagnostics. Compared to general data, we have fewer access to medical cases and high-qualitytraining data. In efforts to enhance the accuracy of the system, we have placed emphasis on appropriate pre-processing techniques, active learning, and transferlearning technology. This article will also highlight the unique computer system used to detect aneurysms in the brain.
目次
Due to advancements in AI-based image processing technology, we are amidst a revolutionary change in its applications in supporting medical image diagnostics. Compared to general data, we have fewer access to medical cases and high-qualitytraining data. In efforts to enhance the accuracy of the system, we have placed emphasis on appropriate pre-processing techniques, active learning, and transferlearning technology. This article will also highlight the unique computer system used to detect aneurysms in the brain.