[The "E-bone" - a one-stop preoperative planning system for reverse total shoulder arthroplasty].

Objective: To develop the'E-Bone', a comprehensive one-stop preoperative planning system for reverse total shoulder arthroplasty with improved accuracy and efficiency.

Methods: The nnU-net deep neural network was utilized for scapula segmentation to obtain precise scapula segmentation results. Based on the 3 key factors, namely bone density, upward and downward angle and nail length, the base was automatically positioned. The quantitative parameters required for surgical planning were calculated. A personalized guide plate was generated by combining glenoid morphology and base positioning information. The system interface was developed to modularize various functions for easy use, providing interactive operation and real-time display.

Results: Compared with the Mimics system, the'E-bone'preoperative planning system reduced complex manual adjustments during the planning process. The average planned nail length was longer than that of the Mimics system, and the planning time was reduced by 86%. The scapula segmentation accuracy of this system reached 99.93%, better than that of Mimics to achieve a higher precision.

Conclusion: The"E-bone"system provides a one-stop, efficient, and accurate preoperative planning system for reverse shoulder replacement and potentially broader clinical applications.