Abstract: Objective  To explore the efficacy of a self-designed intelligent robot-assisted reduction system in the closed reduction of femoral shaft fracture by a cadaveric study. Methods  Four fresh frozen cadavers (2 male and 2 female ones) were used in this study, and 8 femoral shaft fractures were created in accordance with clinical cases (AO classification, 1 case of type A1, 3 cases of type A2, 2 cases of type A3 and 2 cases of type B2). A self-designed intelligent robot-assisted reduction system was used to assist the closed reduction in the cadaveric models. Intraoperative registration time, autonomous reduction time and total operation time were recorded. The length and femoral neck anteversion before osteotomy and after reduction were measured, and the three-dimensional model of the femur before operation and after reduction was matched using Geomagic Qualify 2013 software, and the reduction error was measured to evaluate the fracture reduction accuracy and quality. Results  Effective reduction was accomplished in 8 bone models with assistance of the self-designed intelligent robot-assisted reduction system. The average intraoperative registration time was （33.3±3.7）min, the autonomous reduction time was (34.8±3.3) min, and the average total operative time was (68.0±4.1) min. The difference in length before osteotomy and after reduction was (1.6±1.6) mm, and the difference in femoral neck anteversion was (4.1±3.5)°. The mean reduction error was (4.3±2.2) mm. Conclusions  Our self-designed intelligent robot-assisted reduction system can assist in completing the closed reduction of femoral shaft fracture, with high reduction accuracy and operational feasibility, reducing the reduction time, avoiding the damage of blood supply around the fracture site, and reducing the exposure to radiation during the operation, it provides a better method for the closed reduction of femoral shaft fracture.

Key words: robots, femoral shaft fracture, closed reduction, anatomy