Abstract

Over 8000 total hip arthroplasties (THA) in the UK were revised in 2019, half for aseptic loosening. It is believed that Artificial Intelligence (AI) could identify or predict failing THA and result in early recognition of poorly performing implants and reduce patient suffering.

The aim of this study is to investigate whether Artificial Intelligence based machine learning (ML) / Deep Learning (DL) techniques can train an algorithm to identify and/or predict failing uncemented THA.

Consent was sought from patients followed up in a single design, uncemented THA implant surveillance study (2010–2021). Oxford hip scores and radiographs were collected at yearly intervals. Radiographs were analysed by 3 observers for presence of markers of implant loosening/failure: periprosthetic lucency, cortical hypertrophy, and pedestal formation.

DL using the RGB ResNet 18 model, with images entered chronologically, was trained according to revision status and radiographic features. Data augmentation and cross validation were used to increase the available training data, reduce bias, and improve verification of results.

184 patients consented to inclusion. 6 (3.2%) patients were revised for aseptic loosening. 2097 radiographs were analysed: 21 (11.4%) patients had three radiographic features of failure.

166 patients were used for ML algorithm testing of 3 scenarios to detect those who were revised. 1) The use of revision as an end point was associated with increased variability in accuracy. The area under the curve (AUC) was 23–97%. 2) Using 2/3 radiographic features associated with failure was associated with improved results, AUC: 75–100%. 3) Using 3/3 radiographic features, had less variability, reduced AUC of 73%, but 5/6 patients who had been revised were identified (total 66 identified).

The best algorithm identified the greatest number of revised hips (5/6), predicting failure 2–8 years before revision, before all radiographic features were visible and before a significant fall in the Oxford Hip score. True-Positive: 0.77, False Positive: 0.29.

ML algorithms can identify failing THA before visible features on radiographs or before PROM scores deteriorate. This is an important finding that could identify failing THA early.

Purpose of the study: Acetabular version is a most important parameter for repositioning the acetabular fragment during periacetabular osteotomy. Recently, a few studies have presented a significant number of dysplastic hips with acetabular retroversion. There have not however been any studies devoted specifically to the severity of bilateral acetabular retroversion. The purpose of this work was to determine the incidence of bilateral retroversion in patients undergoing periacetabular osteotomy for dysplasia in order to identify and validate a retroversion index which would be predictive of the degree of retroversion. This index could be added to congenital hip dysplasia classifications to include acetabular version.

Material and methods: The Lequesne lateral view of the hip was obtained in 174 patients (348 hips, 137 women and 37 men, mean age 30 years) undergoing periacetabular osteotomy for symptomatic dysplasia. One hundred ninety-five hips (56%) were operated on and 153 (44%) were considered normal or non-symptomatic and were not operated. The following parameters were noted for each hip: VCE, VCA, HTE, femoral head extrusion, index of acetabular depth, crossing-over, retroversion index. The retroversion index was checked on a bone model of the pelvis which was x-rayed in the neutral position then turned progressively. Statistical data were analyzed with SAS.

Results: Five percent of the operated hips presented neutral version, 53% anteversion and 42% retroversion. Twenty-four percent of the non-operated hips were normal, 22% presented pure retroversion and 54% were dysplastic. All of the measurements were significantly deviated towards dysplasia for operated hips, with the exception of the retroversion index and the VCA.

Discussion: These data validated the retroversion index and confirmed that one out of three dysplastic hips displays retroversion. In addition, it would appear that for dysplastic hips with retroversion, the degree of lateral coverage or the HTE angle determines whether surgery is needed or not and not the degree of retroversion. But as pure retroversion can be symptomatic in itself, and since the majority of these version or cover anomalies can be treated by periacetabular osteotomy, we propose a classification of hip dysplasia included acetabular version.

Conclusion: This classification is designed as an aid for the orthopedic surgeon for reorienting the acetabular fragment to obtain the optimal position.

Purpose of the study: The majority of acetabular bone defects observed during revision hip surgery can be treated with a hemispheric implant, associated or not with a bone graft. In many patients however, loss of bone stock is so great that a more complex system must be used with a sustaining ring, multilobulated implants, or massive allografts. All have their technical difficulties or problems with fixation. The purpose of this work was to evaluate a new technique for acetabular reconstruction using modular implants fashioned with a new biomaterial, porous tantalum, which had specific properties favoring osteointegration.

Material and methods: These modular implants were fashioned so as to enable reconstruction of the acetabular cavity in cases with complex loss of bone stock. The design allows simultaneous biological incorporation and mechanical support with a press-fit hemispheric cup. These implants were used for 16 hips (16 patients, 12 women and 4 men, mean age 63.6 years, age range 34–86 years). These patients were followed for 31.9 months on average (range 24–39 months). The acetabular defects were Paprosky 2A (n=1), 2B ‘n=3), 2C (n=1), 3A (n=5), 3B (n=6). On average, these patients had undergone 2.8 cup replacements (1–9) on the same hip.

Results: The mean Harris hip score improved from 39.31 (range 33–52) preoperatively to 75.18 (range 52–92) at last follow-up. Preoperatively, the center of rotation of the prosthetic hip was situated a a mean horizontal distance of 18.6 mm (range −3 to 46 mm) and a mean vertical distance of 27.6 mm (range −16 to 52 mm) from the ideal center of rotation according to Ranawat. Postoperatively, the prosthetic center of rotation was situated at a mean horizontal distance of 10.5 mm (range 1–25 mm) and a mean vertical distance of 7.4 mm (range −15 to 25 mm) front the ideal center of rotation. None of the implants presented loosening or migration at last follow-up.

Discussion: At short-term follow-up, this modular system for acetabular reconstruction has provided good results for acetabular reconstruction which can accept a hemispheric cup alone and which would have required use of other reconstruction methods such as structural allografts, sustaining rings or other.

Conclusion: A longer follow-up will be needed to determine whether these good clinical and radiological results persist with time.