JOA - 2026-05-08 - Journal Article
Accuracy of Fluoroscopic Navigation for Predicting Leg Length and Offset in Primary Hip Arthroplasty.
Murphy MP, Kelly P, Dunn JR, Hamilton G, Clift A, Hamilton WG
Topics
Key Takeaway
Fluoroscopic navigation predicted leg length within ±5 mm in 86.0% of cases but femoral offset within ±5 mm in only 64.9% of cases, with navigation systematically underestimating leg length change (mean 2.5 vs 4.1 mm actual) and overestimating offset change (mean 3.4 vs 1.7 mm actual).
Summary Depth
Choose how much analysis to show on this article page.
Summary
This study evaluated the accuracy of a fluoroscopy-based intraoperative navigation system for predicting leg length and femoral offset changes in 490 consecutive primary THAs performed by three surgeons at a single academic center. Intraoperative navigation predictions were compared to AI-based measurements on pre- and postoperative AP pelvis radiographs. Navigation systematically underestimated leg length change (2.5 vs 4.1 mm, P<0.0001) and overestimated offset change (3.4 vs 1.7 mm, P<0.0001), with offset accuracy within ±5 mm in only 64.9% of cases versus 86.0% for leg length.
Key Limitation
The postoperative radiographic timepoint is not specified, meaning early postoperative swelling, patient positioning variability, and pelvic tilt changes between pre- and postoperative films may introduce measurement error into the AI-based reference standard itself.
Original Abstract
BACKGROUND
Fluoroscopic-assisted computer navigation is widely used to guide intraoperative decisions on leg length (LL) and offset in primary total hip arthroplasty (THA), but its accuracy remains under-evaluated. This study assessed the precision of a fluoroscopy-based navigation system by comparing its intraoperative predictions to validated postoperative measurements.
METHODS
We retrospectively reviewed 490 consecutive primary THA cases performed by three surgeons at a tertiary academic center. Intraoperative changes in leg length and offset using a fluoroscopic navigation system were recorded. Pre- and postoperative antero-posterior pelvis radiographs were measured using an artificial intelligence (AI)-based algorithm to determine actual changes in LL and offset. Navigated and radiographic values were compared using mean/median differences and by percentage of cases within clinically relevant thresholds: ± three mm, ± five mm, and ± 10 mm.
RESULTS
For leg length, postoperative AI-based measurements showed a mean change of 4.1 ± 5.6 mm (median 3.7), while the intraoperative navigation system predicted a mean change of 2.5 ± 3.1 mm (median two, P < 0.0001). Navigation underestimated leg length change on average. Predictions were within ± three mm in 65.8% of cases, ± five mm in 86.0%, and ± 10 mm in 98.5%.For femoral offset, AI-based measurements showed a mean postoperative change of 1.7 ± 5.4 mm (median 1.6), compared with a mean predicted change of 3.4 ± 5.0 mm (median three) from navigation (P < 0.0001). Navigation overestimated offset change on average. Accuracy was within ± three mm in 44.1% of cases, ± five mm in 64.9%, and ± 10 mm in 93.9%.
CONCLUSION
Although the fluoroscopy-based navigation system showed statistically significant differences from postoperative radiographic measurements, it achieved clinically acceptable accuracy in most cases. These findings support the system's utility in guiding THA while emphasizing the importance of intraoperative awareness of its limitations.