Multi-source information fusion using CNN-LSTM-Attention for bone layer recognition in robotic orthopedic grinding
Medical & Biological Engineering & Computing · 04.08.2026
DOI: 10.1007/s11517-026-03649-2 ↗Kurz gefasst
Fachartikel in der Zeitschrift Medical & Biological Engineering & Computing, online erschienen am 04.08.2026. Verfasst von Kai Yang und 5 weiteren Autor:innen. Beteiligte Einrichtung: School of Intelligent Engineering and Automation, Beijing University of Posts and Telecommunications, Beijing, 100876, China. Hauptthemen laut MEDLINE: Robotic Surgical Procedures, Orthopedic Procedures, Bone and Bones. Die Arbeit nennt 48 Literaturangaben. Gefördert von: Guizhou Provincial Science and Technology Major Project, Beijing Natural Science Foundation of China, Guizhou Provincial Health and Wellness High-Quality Development Medical Research Joint Fund.
Abstract (Originalsprache)
Epiphyseal opening requires precise localization for bony bridge resection. Traditional surgery is challenged by unclear bony bridge boundary localization and inaccurate grinding precision, while existing robot-assisted operations predominantly focus on pre-operative localization with limited intraoperative autonomous decision-making capabilities. To address this, we propose a multi-source information fusion framework using a CNN-LSTM-Attention network for real-time bone layer differentiation, specifically identifying idling, cancellous, and cortical bone states, during robotic orthopedic grinding. First, the mapping relationship between acceleration, force and acoustic signals and bone density is analyzed, serving as the basis of bone layer recognition. Second, a three-channel parallel late-feature-fusion CNN-LSTM-Attention network is established. The dataset was constructed from multiple independent grinding trials with trial-wise splitting to prevent data leakage under representative robotic grinding conditions. Over five independent runs, the proposed method achieves a test accuracy of 95.06% ± 0.53%, significantly outperforming comparison models including CNN, CNN-Attention, CNN-LSTM, and non-deep-learning baselines (Random Forest, Extra Trees, SVM, KNN). Ablation studies isolating CNN, LSTM, and Attention contributions are provided, and the learned Squeeze-and-Excitation attention weights are visualized to confirm dynamic cross-modal feature weighting. With approximately 0.32 million parameters, the model introduces an inference latency of 2.8 ms on a desktop CPU and 1.2 ms on an NVIDIA Jetson Orin, well below the 50 ms robot control cycle, confirming real-time feasibility. Additionally, we explore single-signal, dual-signal, and triple-signal fusion settings, demonstrating that tri-modal fusion achieves the best performance. Graphical Abstract
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Intelligent Engineering and Automation, Beijing University of Posts and Telecommunications, Beijing, 100876, China.
- School of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing, 100876, China. juxianghuang@bupt.edu.cn.
- School of Intelligent Engineering and Automation, Beijing University of Posts and Telecommunications, Beijing, 100876, China. buptcg@163.com.
- Pediatric Orthopaedics Department, Beijing Jishuitan Hospital, Capital Medical University, Beijing, 100035, China.
- National Regional Medical Center, Beijing Jishuitan Hospital Guizhou Hospital, Guizhou, 550014, China.
Förderung
- Guizhou Provincial Science and Technology Major Project
- Beijing Natural Science Foundation of China
- Guizhou Provincial Health and Wellness High-Quality Development Medical Research Joint Fund
- the Guizhou Provincial Science and Technology Program Project
- Guizhou Provincial Health Commission Science and Technology Fund Project
- Beijing Key Laboratory of Work Safety Intelligent Monitoring
Zitieren
APA
Yang, K., Jia, Q., Huang, J., Chen, G., Feng, C., & Xu, Y. (2026). Multi-source information fusion using CNN-LSTM-Attention for bone layer recognition in robotic orthopedic grinding. Medical & Biological Engineering & Computing. https://doi.org/10.1007/s11517-026-03649-2
Vancouver
Yang K, Jia Q, Huang J, Chen G, Feng C, Xu Y. Multi-source information fusion using CNN-LSTM-Attention for bone layer recognition in robotic orthopedic grinding. Medical & Biological Engineering & Computing. 2026. doi:10.1007/s11517-026-03649-2
BibTeX
@article{Yang2026,
author = {Yang, Kai and Jia, Qingxuan and Huang, Juxiang and Chen, Gang and Feng, Chao and Xu, Yunfeng},
title = {Multi-source information fusion using CNN-LSTM-Attention for bone layer recognition in robotic orthopedic grinding},
journal = {Medical & Biological Engineering & Computing},
year = {2026},
doi = {10.1007/s11517-026-03649-2}
}
EUVIMED