A radiographic clinic for developmental dysplasia of the hip (DDH)

Deep learning-based automated measurement of hip key angles and auxiliary diagnosis of developmental dysplasia of the hip

OBJECTIVES

Anteroposterior pelvic radiographs remains the most widely employed method for diagnosing developmental dysplasia of the hip. This study aims to evaluate the accuracy of an artificial intelligence model in measuring angles in pelvic radiographs of the hip. The assessment seeks to demonstrate the efficacy of the artificial intelligence model in diagnosing both developmental dysplasia of the hip and borderline developmental dysplasia of the hip through the analysis of pelvic radiographs.

METHODS

A total of 1,029 patients, including 273 men and 757 women, were retrospectively included in this study. The anteroposterior pelvic radiographs were randomly divided into three sets: the training set (720 cases), the validation set (103 cases), and the test set (206 cases). Key anatomical points on the anteroposterior pelvic radiographs were identified. The Sharp, Tönnis, and Center Edge angles were calculated automatically based on the corresponding criteria. The hip development status was compared between measurements obtained from the artificial intelligence model and those defined manually by two radiologists. The area under the receiver operating characteristic curve was utilized to assess the diagnostic performance of the artificial intelligence model.

RESULTS

The results obtained from both manual measurements and the artificial intelligence model demonstrated no significant differences in the Sharp, Tönnis, and Center edge angles (all p > 0.05). The intra-class correlation coefficients and correlation coefficient r values exceeded 0.75, indicating that both the artificial intelligence model and manual measurements exhibited good repeatability and a positive correlation. Notably, the artificial intelligence model provided measurements more faster than those conducted by radiologists (p = 0.001). The artificial intelligence model also demonstrated high diagnostic accuracy, sensitivity, and specificity for developmental dysplasia of the hip. The performance of the artificial intelligence model in diagnosing developmental dysplasia of the hip was robust. Additionally, the results from the artificial intelligence model and manual measurements were largely consistent with clinical diagnosis results (p = 0.01). The artificial intelligence model can effectively evaluate hip conditions by measuring the Sharp, Tönnis, and Center edge angles, which are consistent closely with clinical diagnosis results.

CONCLUSIONS

The results of the artificial intelligence model measurements demonstrate a high degree of consistency with those obtained through manual measurements. The angles of Sharp, Tönnis, and Center edge, as evaluated by the deep learning-based convolutional neural network model, exhibit robust diagnostic performance in identifying both developmental dysplasia of the hip and borderline developmental dysplasia of the hip. Consequently, the artificial intelligence model has the potential to fully replace manual measurements of these critical hip angles, providing a more efficient and precise alternative for diagnosing both conditions of the hip.

A Virtual Clinic for DDH Pavlik Harness Monitoring

BACKGROUND

Pavlik harness (PH) treatment is successful in treating over 90% of newborns with developmental dysplasia of the hip (DDH). There is a need for close supervision, frequent adjustments, size changes, and monitoring of complications. This paper aims to provide a safe criterion for remote follow-up of DDH patients treated in a PH to reduce the risk of COVID-19 (coronavirus disease 2019) exposure to patients, parents, and health practitioners.

METHODS

All infants with stable hips (Graf I, IIa/b/c/d, treated III/IV) with consenting parents after appropriate counseling were enrolled in a virtual clinic. Clinics were conducted using the NHS "Attend anywhere" virtual link service by an extended scope practitioner-specialist physiotherapist and a clinical nurse specialist. The virtual clinic group was compared with a matched cohort of patients from 2018/2019.

RESULTS

A total of 141 patients were referred to the neonatal hip clinic; 45 patients were eligible for harness treatment and 20 patients were selected for virtual clinics. In total, there were 35 virtual clinic appointments. Each of the patients had an average of 1.7 virtual appointments ranging from 1 to 3 (26.3% of total number of clinics). Age at presentation of the treated group was 7±4.2 weeks and control group 5.7±5.5 weeks (P=0.59). PH duration of the study group was 9±2.6 weeks and the control group, 7.8±2.5 weeks (P=0.12). There were no missed complications at the follow-up face-to-face appointment. Patients saved an average of 76 km total travel distance.

CONCLUSIONS

This study demonstrates adequate evidence that children requiring routine follow-up appointments involving PH adjustment, skincare, and identification of clinical anomalies, can be treated and followed up safely using virtual clinics. Clinical triage of suitable patients for virtual clinic provision must always be made by experienced clinicians. Children presenting with Graf IIa, IIb, IIc, IId, as well as those with stable and improving Graf III at initial diagnoses, had successful treatment with virtual clinic follow-up appointments in this study.

LEVEL OF EVIDENCE

Level IV.

Knee Flexion Angle Measurement Using Virtual Assessment Tools: Correct Procedure and Potential Pitfalls

Virtual patient assessment will inevitably require smartphone technology to remotely measure knee range of motion. We conducted an experiment to analyze the impact of observer position relative to the flexed knee on the perceived angle measured using an electronic application (Dr. Goniometer) for iPhone. Two observers measured the apparent knee flexion angle from 7 different positions at 3 different heights relative to the center of the knee joint. Intraclass correlations were calculated to evaluate the intraobserver and interobserver variability using two-way mixed-effects models. The intraclass correlation for interobserver variability was excellent at 0.804 (95% confidence interval 0.663-0.889). When the observer was greater than 15° from the knee perpendicular, the true angle of knee flexion (90°) was not observed in any of the measurements. This was the case when observed from both proximal (range 95°-121°) and distal (range 92°-108°) directions. Ideally the camera lens should be perpendicular to the long axis of the lower limb in the proximal-distal direction and at the same height. However, if the camera lens is within 15° of the perpendicular, then at 90° of true flexion, the perceived angle will not be greater than 95° in 94% of cases.

Establishing a Virtual Clinic for Developmental Dysplasia of the Hip: A Prospective Study

BACKGROUND

The aim was to describe the introduction and operation of a virtual developmental dysplasia of the hip (DDH) clinic. Our secondary objectives were to provide an overview of DDH referral reasons, treatment outcomes, and adverse events associated with it.

METHODS

A prospective observational study involving all patients referred to the virtual DDH clinic was conducted. The clinic consultant delivered with 2 DDH clinical nurse specialists (CNS). The outcomes following virtual review include further virtual review, CNS review, consultant review or discharge. Treatment options include surveillance, brace therapy, or surgery. Efficiency and cost analysis were assessed.

RESULTS

Over the 3.5-year study period, 1002 patients were reviewed, of which 743 (74.2%) were female. The median age at time of referral was 7 months, (interquartile range of 5 to 11) with a median time to treatment decision of 9 days. Median waiting times from referral to treatment decision was reduced by over 70%. There were 639 virtual reviews, 186 CNS reviews, and 144 consultant reviews. The direct discharge rate was 24%. One hundred one patients (10%) had dislocated or subluxed hips at initial visit while 26.3% had radiographically normal hips. Over the study period 704 face to face (F2F) visits were avoided. Cost reductions of €170 were achieved per patient, with €588,804 achieved in total. Eighteen parents (1.8%) opted for F2F instead of virtual review. There were no unscheduled rereferrals or recorded adverse events.

CONCLUSION

We report the outcomes of the first prospective virtual DDH clinic. This clinic has demonstrated efficiency and cost-effectiveness, without reported adverse outcomes to date. It is an option to provide consultant delivered DDH care, while reducing F2F consults.

LEVEL OF EVIDENCE

Level III.