← 研究テーマ一覧
Domain adaptation & generalization

環境変化に強い医療AI

施設、撮影装置、標本の状態が変わっても使えるように、医用画像のドメイン適応と汎化を研究します。

Selected Papers

代表論文

ソースとターゲット領域をランキング情報で対応づける図
Fig. 1 · ランキングを利用したドメイン適応 · 出典：論文PDF ↗
ISBI 2026 · 半教師ありドメイン適応 · 重症度の順序

Ranking-Guided Semi-Supervised Domain Adaptation for Severity Classification

順位関係を手掛かりに、施設差のある内視鏡画像で重症度を推定します。

Fig. 2 · ヒートマップによる擬似ラベル
Fig. 2 · ヒートマップによる擬似ラベル · 出典：論文PDF ↗
Selected Medical Image Analysis 2022 · 教師なしドメイン適応

Effective pseudo-labeling based on heatmap for unsupervised domain adaptation in cell detection

擬似的な細胞位置ヒートマップを利用し、撮影条件が異なる細胞画像の検出器を適応させます。

論文・アブストラクトを見る ↗
Fig. 2 · 細胞位置ヒートマップの反復学習
Fig. 2 · 細胞位置ヒートマップの反復学習 · 出典：論文PDF ↗
MICCAI 2021 · ドメインシフト下の細胞検出

Cell Detection in Domain Shift Problem Using Pseudo-Cell-Position Heatmap

ソースとターゲットの細胞画像の違いを、擬似細胞位置ヒートマップと反復学習で扱います。

論文・アブストラクトを見る ↗
Publications

関連論文

  1. Ranking-Guided Semi-Supervised Domain Adaptation for Severity Classification
    Shota Harada, Ryoma Bise, Kiyohito Tanaka, and Seiichi Uchida · IEEE 23rd International Symposium on Biomedical Imaging (ISBI), 2026.
  2. Domain Adaptation for Ulcerative Colitis Severity Estimation Using Patient-Level Diagnoses
    T Yamaguchi, BK Iwana, R Bise, S Harada, T Okuo, K Tanaka, K Shiku · Workshop on MICCAI: MLMI, 2025.
  3. Domain Generalization of Pathological Image Segmentation by Patch-Level and WSI-Level Contrastive Learning
    Yuki Shigeyasu, Shota Harada, Akihiko Yoshizawa, Kazuhiro Terada, Naoki Nakazima, Mariyo Kurata, Hiroyuki Abe, Tetsuo Ushiku, Ryoma Bise · International Conference on Machine Vision Applications (MVA), 2025
  4. Weakly-Supervised Domain Adaptation with Proportion-Constrained Pseudo-Labeling
    Takumi Okuo, Shinnosuke Matsuo, Shota Harada, Kiyohito Tanaka, and Ryoma Bise · International Joint Conference on Neural Networks (IJCNN), 2025
  5. Domain Generalization for Pathological Images Using the Storage Period Information
    Yuki Shigeyasu, Shota Harada, Akihiko Yoshizawa, Kazuhiro Terada, Naoki Nakazima, Mariyo Kurata, Hiroyuki Abe, Tetsuo Ushiku, and Ryoma Bise · IEEE International Symposium on Biomedical Imaging (ISBI), 2024.
  6. Cluster Entropy: Active Domain Adaptation in Pathological Image Segmentation
    Xiaoqing Liu, Kengo Araki, Shota Harada, Akihiko Yoshizawa, Kazuhiro Terada, Mariyo Kurata, Naoki Nakajima, Hiroyuki Abe, Tetsuo Ushiku, and Ryoma Bise · IEEE International Symposium on Biomedical Imaging (ISBI), 2023. (Oral)
  7. Cluster-Guided Semi-Supervised Domain Adaptation for Imbalanced Medical Image Classification
    Shota Harada, Ryoma Bise, Kengo Araki, Akihiko Yoshizawa, Kazuhiro Terada, Mariyo Kurata, Naoki Nakajima, Hiroyuki Abe, Tetsuo Ushiku, and Seiichi Uchida · IEEE International Symposium on Biomedical Imaging (ISBI), 2023.
  8. Effective pseudo-labeling based on heatmap for unsupervised domain adaptation in cell detection
    Hyeonwoo Cho, Kazuya Nishimura, Kazuhide Watanabe, and Ryoma Bise · Medical Image Analysis, vol.79, 102436, https://doi.org/10.1016/j.media.2022.102436, 2022 (top journal in medicalImage analysis, IF:13.828)
  9. Cell Detection in Domain Shift Problem Using Pseudo-Cell-Position Heatmap
    Hyeonwoo Cho, Kazuya Nishimura, Kazuhide Watanabe, and Ryoma Bise · International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI2021), 2021, (accepted, top conference in medial image analysis, Provisional acceptanc

Publication一覧を見る →