| 講演抄録/キーワード |
| 講演名 |
2025-08-28 15:20
Muscle region segmentation for adhesion analysis in shoulder ultrasound monitoring system
-- Supervised segmentation using acoustic morphology and superpixel-based feature fusion -- ○Jinyi Zhao(Science Tokyo)・Mizuki Fujiwara(JR Sendai Hospital, Tohoku Univ.)・Masahiro Yamaguchi・Marie Tabaru(Science Tokyo) US2025-30 |
| 抄録 |
(和) |
Shoulder muscle adhesions, especially between the deltoid and subscapularis, contribute significantly to mobility impairments. While prior studies applied speckle tracking to manually defined ROIs, such methods are time-consuming and operator-dependent. We propose a fully automated system that segments shoulder ultrasound images and extracts the deltoid-subscapularis boundary using acoustic morphology and superpixel-based feature fusion. Key tissue regions are identified through multi-stage clustering and refined via contour-based boundary extraction. Preliminary results show strong agreement between extracted boundaries and anatomical expectations. This approach removes manual intervention, improves consistency, and lays the foundation for future integration of motion analysis techniques such as optical flow in clinical assessments. |
| (英) |
Shoulder muscle adhesions, especially between the deltoid and subscapularis, contribute significantly to mobility impairments. While prior studies applied speckle tracking to manually defined ROIs, such methods are time-consuming and operator-dependent. We propose a fully automated system that segments shoulder ultrasound images and extracts the deltoid-subscapularis boundary using acoustic morphology and superpixel-based feature fusion. Key tissue regions are identified through multi-stage clustering and refined via contour-based boundary extraction. Preliminary results show strong agreement between extracted boundaries and anatomical expectations. This approach removes manual intervention, improves consistency, and lays the foundation for future integration of motion analysis techniques such as optical flow in clinical assessments. |
| キーワード |
(和) |
shoulder ultrasound image / region segmentation / clustering algorithm / superpixel segmentation / / / / |
| (英) |
shoulder ultrasound image / region segmentation / clustering algorithm / superpixel segmentation / / / / |
| 文献情報 |
信学技報, vol. 125, no. 159, US2025-30, pp. 26-30, 2025年8月. |
| 資料番号 |
US2025-30 |
| 発行日 |
2025-08-21 (US) |
| ISSN |
Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
| PDFダウンロード |
US2025-30 |
|