Abstract
Dynamic 2D sonography of the infant hip is a commonly used clinical procedure for developmental dysplasia of the hip (DDH) screening. It however has been found to be unreliable with some studies reporting associated misdiagnosis rates of up to 29%. In a recent systematic review, Charlton et al. examined dynamic ultrasound (US) screening for hip instability in the first six weeks after birth and found current best practices for such early screening techniques to be divergent between international institutions in terms of clinical scanning protocols. Such protocols include: the appropriate scanning plane and US probe position (e.g. coronal, transverse, lateral, anterior), DDH diagnostic metrics (e.g. femoral head coverage, alpha angle), appropriate patient age when scanning, and follow up procedures. To improve reliability of diagnosis and to help in standardizing diagnosis across different raters and health-centers, we propose an automated method for dynamically assessing hip instability using 3D US.
38 infant hips from 19 patients were scanned with B-mode 3D US by a paediatric orthopaedic surgeon and two technologists from the radiology department at a paediatric tertiary care centre. To quantify hip assessment, we proposed the use of femoral head coverage variability (ΔFHC3D) within 3D US volumes collected during a sequence of US scans (one at rest, and another with posterior stress applied to the joint as maneuvered during a dynamic assessment). We used phase symmetry image features to localize the ilium's vertical cortex and a random forest classifier to identify the location of the femoral head.
The proposed ΔFHC3D provided good repeatability with an average test-retest ICC measure of 0.70 (95% confidence interval: 0.35 to 0.87, F(21,21) = 7.738, p<.001). The mean difference of ΔFHC3D measurements was 0.61% with a SD of 4.05%.
Since the observed changes in ΔFHC3D start near 0% and range up to about 18% from stable to mildly unstable hips in this cohort, the mean difference and standard deviation of ΔFHC3D measurements observed suggest that the proposed metric and technique likely have sufficient resolution and repeatability to quantify differences in hip laxity. The long-term significance of this approach to evaluating dynamic assessments may lie in increasing early diagnostic accuracy in order to prevent dysplasia remaining undetected prior to manifesting itself in early adulthood joint disease.
