Work

AI-assisted direct anterior approach versus posterolateral approach in total hip arthroplasty: a retrospective cohort study based on artifact-reduced CT 3D reconstruction

Year: 2025

Type: article

Abstract: Objective: To compare the accuracy of implant positioning and early functional recovery between direct anterior approach (DAA) and posterolateral approach (PLA) in total hip arthroplasty (THA) guided by an artificial intelligence preoperative planning system (AIHIP). Methods: The study population consisted of 206 patients who underwent DAA surgery and 81 patients who underwent PLA surgery, all of whom were designed preoperatively using AI-HIP, and postoperatively using artefact-reduced CT reconstruction for prosthesis mounting angle measurements and follow-up such as postoperative outcomes. The main assessments included prosthesis positioning accuracy (compared to the preoperative plan): acetabular anterior inclination (AA), femur anterior inclination (FNA), combined anterior inclination (CA), alignment of femoral stem prosthesis and femur; clinical outcomes: operative time, hospital stay, and time to grounding; functional scores: Harris Hip Score, WOMAC, and VAS Pain Score; and biomarkers: haemoglobin, CRP, and IL-6, among others. Results: All 287 patients completed ≥6-month follow-up. While preoperative femoral/acetabular anteversion showed no intergroup differences (p > 0.05), the direct anterior approach (DAA) demonstrated superior postoperative acetabular anteversion control (20.93 ± 7.54° vs. 24.34 ± 7.93°, p < 0.001) despite comparable femoral anteversion (12.97 ± 6.93° vs. 14.56 ± 7.21°, p = 0.009). AI-assisted predictions exhibited smaller deviations in DAA for both parameters (FNA: 3.12 ± 5.88° vs. 5.59 ± 8.21°, p = 0.005; AA: 0.93 ± 7.54° vs. -4.34 ± 7.93°, p < 0.001). No significant differences emerged in combined anteversion, acetabular abduction, or femoral stem alignment parameters (all p > 0.05). DAA achieved shorter incisions (10.64 ± 0.94 vs. 15.21 ± 1.33 cm, p < 0.001) and hospital stays (7.59 ± 4.18 vs. 9.09 ± 3.65 days, p < 0.001) despite longer operative times (118.67 ± 26.95 vs. 53.27 ± 58.71 min, p < 0.001). Functional recovery favored DAA, with better VAS/Harris scores at 3 months and WOMAC scores at 1 month (all p < 0.05). No intergroup differences were observed in postoperative CK, CRP, Hb, or IL-6 levels (p > 0.05). Conclusion: Both DAA and PLA approaches resulted in satisfactory postoperative outcomes; however, the DAA approach demonstrated enhanced early postoperative efficacy indicators, as well as improved femoral neck and acetabular anteversion compared to the PLA approach. This study advocates for the preferential adoption of the DAA technique for THA, while also emphasizing the importance of considering individual patient factors, as well as the surgeon's preferences and expertise.

Source: Frontiers in Bioengineering and Biotechnology

DOI: 10.3389/fbioe.2025.1509200