Myles Mabee BEng, Presenter: Nothing to Disclose

Jacob L. Jaremko MD, PhD, Abstract Co-Author: Nothing to Disclose

Richard Thompson PhD, Abstract Co-Author: Nothing to Disclose

Accuracy of 2D ultrasound detection of developmental dysplasia of the hip (DDH) is limited by variation in acetabular appearance and alpha angles with changes in ultrasound probe orientation. 3D ultrasound can capture the entire acetabular shape in ~3 seconds. Reproducible identification of a “standard plane” from landmarks within 3D ultrasound would permit measurement of the alpha angle and other indices in that plane on each scan, reducing inter-scan variability vs. traditional 2D ultrasound while providing a more representative image of the hip. We noted that the anterior and posterior edges of the acetabular rim are relatively constant landmarks which may be reproducibly identifiable. We determined intra- and inter- observer variability of identification of these two landmarks, of the orientation of the resulting “standard plane,” and of the alpha angle measured in this plane.

We performed 2D and 3D ultrasound scans of 51 hips in 42 patients, 20 normal, 10 borderline initially but normalizing at follow up, and 12 treated for DDH. Two users each identified 2 landmarks within each 3D scan, blinded to each other’s findings and clinical data, and repeated this after a 1-week delay. For each user at each session, we recorded 3D landmark locations, angles between the resulting “standard plane” orientations, and reliability of acetabular alpha angles measured on 3D “standard plane” images vs. concurrently acquired 2D scans.

The 3D distances between user-identified landmarks were 0.9+/-0.9 mm (mean +/- standard deviation SD) intraobserver, 1.6+/-1.0 mm between users. Angles between “standard planes” calculated from these landmarks varied by 4.2+/-3.9° intraobserver and 5.4+/-4.9° between users. Comparison of 3D and 2D alpha angles yielded a difference of -0.5+/-6.7°. All parameters trended toward lower variability in normal hips than dysplastic hips.

Acetabular landmarks and a “standard plane” calculated from these can be identified with high reliability within hip 3D ultrasound. Measuring indices in this standard plane has potential to improve reliability and accuracy of ultrasound assessment of DDH.

A “standard plane” extracted from 3D ultrasound may more reproducibly capture infant hip shape than current 2D ultrasound, potentially improving accuracy and reliability of hip dysplasia diagnosis.

Mabee, M, Jaremko, J, Thompson, R, How Reproducibly Can Landmarks for a Standardized Coordinate System be Obtained from 3D Hip Ultrasound?.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14011445.html