IOA 2022

Development of Reliable Fully Automated Surface Topographic Measurements to Assess Shoulder Asymmetry in Adolescent Idiopathic Scoliosis Patients

Roger F. Widmann 1 Ankush Thakur 1 Benjamin Groisser 2 Howard Hillstrom 3 Matthew Cunningham 1 Michael T. Hresko 5 Hila Otremski 6 Kyle W. Morse 1 Kira Page 1 Ron Kimmel 4 Alon Wolf 2
1Department of Orthopedics, Hospital for Special Surgery, USA
2Department of Mechanical Engineering, Technion-Israel Institute of Technology, Israel
3Department of Rehabilitation, Hospital for Special Surgery, USA
4Department of Computer Science, Technion-Israel Institute of Technology, Israel
5Department of Orthopedic Surgery, Boston Children's Hospital, USA
6Department of Orthopedic Surgery, Dana Dwek Children's Hospital, Israel

Shoulder asymmetry is an important concern for adolescent idiopathic scoliosis (AIS) patients, but there is a need for objective, validated measurements to describe this aspect of trunk deformity. Traditional radiographic and clinical evaluations (e.g. acromion height) correlate poorly with patient reported outcomes. The purpose of this study is to develop novel surface topographic (ST) tools for objective quantification of shoulder asymmetry measurements and to investigate correlations with the Trunk Appearance Perception Scale (TAPS).

Topographic scans were captured with a 3dMD whole-body surface scanner on a cohort of 92 patients with Adolescent Idiopathic Scoliosis (AIS) and 27 controls ages 11 to 20 years. All subjects underwent standardized clinical assessment and completed TAPS questionnaires, while a subset (N=46) underwent repeated scanning with multiple raters for reliability analysis. The AIS group also completed whole-body EOS biplanar imaging studies (Cobb angle range 6.39 to 78.94 degrees). A markerless automated pipeline was used to compute a series of volumetric and angular measures of shoulder asymmetry. Intraclass correlations (ICCs) were computed for each of the ST measures, and linear regressions were fitted to model correlation with TAPS.

Shoulder ST measurements exhibited good to excellent reliability with inter-rater ICCs ranging from 0.75 to 0.9 (Table 1). Shoulder asymmetry topographic measures were all significantly correlated with TAPS (p<0.01): Acromia height angle (R=0.26), Shoulder volume asymmetry (R=0.30), and shoulder normal asymmetry (R=0.40).

Shoulder ST measurements are reliable, and demonstrate significant correlation with TAPS. This work presents an automated, objective, reliable tool for surface and volumetric evaluation of shoulder balance. These measurements may provide valuable objective data for the management of AIS patients.