AI-Driven Real-Time Decision Support in Arthroscopic Procedures Using Computer Vision
DOI:
https://doi.org/10.63282/3050-9416.IJAIBDCMS-V7I2P116Keywords:
Artificial Intelligence, Computer Vision, Arthroscopy, Real-Time Systems, Surgical Decision Support, Deep Learning, Minimally Invasive SurgeryAbstract
The clinical efficacy of arthroscopic repair remains heavily dependent on a surgeon's spatial orientation and real-time interpretation of constrained visual fields. While intraoperative assistance is evolving, current platforms often struggle with high-latency processing and poor anatomical differentiation. We developed a high-performance computer vision framework that synchronizes anatomical segmentation with instrument tracking to provide instantaneous surgical guidance. By deploying hybrid CNN-Transformer architecture via edge computing, the system achieves sub-30ms latency meeting the strict requirements for fluid, real-time feedback in the operating room. Our evaluation across multi-institutional datasets shows a significant reduction in procedural deviation and improved accuracy in identifying critical structures like ligaments and meniscal boundaries.
References
1. Buchanan et al., “Surgical computer vision for intraoperative decision-support,” Artificial Intelligence Surgery, 2026.
2. Watson et al., “Real-time AI in operating room analytics,” PMC, 2025.
3. Guo et al., “Applications of computer vision in laparoscopic surgery,” ScienceDirect, 2023.
4. Protserov et al., “Real-time surgical guidance using deep learning,” Nature Digital Medicine, 2024.
5. Shu et al., “DualVision ArthroNav,” arXiv, 2025.
