Development and evaluation of deep learning models for detecting and classifying various bone tumours in full-field limb radiographs using automated object detection models
The DINO model automatically detects bone tumours better than the YOLO model, and may assist doctors in detecting tumours and classifying malignant/benign bone tumours in clinical practice.
Authors
- Yasuharu Nakashima (ORCID: https://orcid.org/0000-0002-6692-982X)
- Nobuhiko Yokoyama (ORCID: https://orcid.org/0000-0002-3013-3205)
- Tomoya Matsunobu (ORCID: https://orcid.org/0000-0002-5431-4975)
- Makoto Endo (ORCID: https://orcid.org/0000-0002-7956-2190)
- Seiichi Uchida (ORCID: https://orcid.org/0000-0001-8592-7566)
- Ryoma Bise (ORCID: https://orcid.org/0000-0002-2214-6719)
- Nokitaka Setsu (ORCID: https://orcid.org/0000-0003-0042-3321)
- Michiko Yamana
Institutions
- Kyushu University (JP)
- National Hospital Organization Kyushu Cancer Center (JP)
- Kyushu Rosai Hospital (JP)
Publication Details
- Journal
- Bone and Joint Research
- Published
- 2025-09-02
- DOI
- https://doi.org/10.1302/2046-3758.149.bjr-2024-0505.r1
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Japan Society for the Promotion of Science