ARVIS® Shoulder

The ARVIS® Shoulder is an advanced surgical navigation system intended to enhance the accuracy of total shoulder arthroplasty by assisting surgeons with intraoperative measurements and implant positioning. It achieves this through a combination of CT-based preoperative planning, utilizing AI-driven image segmentation and landmarking algorithms, and real-time intraoperative tracking of instruments relative to the patient's anatomy.

The ARVIS® Shoulder is a computer-controlled surgical navigation system designed to provide intra-operative measurements to surgeons, aiding in the selection and positioning of orthopedic implant components specifically for total shoulder arthroplasty. This system integrates software, electronic hardware worn by the surgeon on the head and waist, and surgical instruments to intraoperatively track tools and locate anatomical structures. The workflow relies on preoperative CT-based reconstruction of the patient's shoulder anatomy and a detailed preoperative plan, which then guides the surgeon in navigating and positioning shoulder instruments and implants, specifically for use with the Enovis AltiVate implant system. The system's preoperative planning platform employs AI-based automatic image segmentation and landmarking algorithms. These algorithms were trained using a dataset of 240 CT scans and validated on a separate test dataset of 60 CT scans from patients undergoing shoulder joint replacement. Non-clinical performance testing, including positional accuracy (ASTM F2554), accuracy after mechanical disturbances (ASTM F3107), benchtop accuracy, cadaveric accuracy, and cadaver validation, was conducted. The software, particularly the AI algorithms, underwent verification and validation, demonstrating that the automatic segmentation (for scapula and humerus) and landmarking (for glenoid center, trigonum, inferior point, medial/lateral epicondyles, and neck plane) met predefined acceptance criteria based on metrics like Dice Similarity Coefficient, Mean Absolute Distance, Hausdorff Distance, Euclidean Distance, and Angular Separation.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the performance of the AI-based automatic image segmentation and landmarking algorithms vary with different CT scanner manufacturers and imaging protocols not explicitly included in the training/testing datasets?; What is the reported accuracy and precision of implant placement when using the ARVIS® Shoulder system compared to traditional techniques or other navigation systems in a clinical setting, beyond the benchtop and cadaveric studies?; Are there any specific patient anatomical variations or pathological conditions of the shoulder (e.g., severe bone loss, previous fractures) that might limit the accuracy or applicability of the ARVIS® Shoulder system's navigation and AI algorithms?; What is the learning curve for surgeons to become proficient with the ARVIS® Shoulder system, particularly regarding the preoperative planning and intraoperative navigation components?; Given its reliance on the Enovis AltiVate implant system, what are the technical challenges and validation requirements for potentially expanding compatibility with other shoulder implant systems in the future?