Combining ultrasound and surface topography on curve assessment versus radiographic measurements for children with adolescent idiopathic scoliosis- a pilot study.

PURPOSE

To determine the correlation in combining ultrasound (US) and surface topography (ST) parameters to assess curve severity for children with adolescent idiopathic scoliosis (AIS).

METHODS

Seventy participants (56 F & 14 M, aged 14.2 ± 1.8 years old), all having (1) radiographs (XR), (2) US scans, and (3) 3D ST on the back were recruited. The reliability of ST measurements was evaluated on the first 30 participants. Two raters (R1, R2) with no prior experience measured ST data twice with one week apart. Then, all the US, ST, and XR images were measured once by R1, who had 10 years of experience in clinical radiology with two years of experience in scoliosis research. The correlations between parameters and stepwise linear regression were used to estimate XR measurements from US and ST. The main statistical measures were mean absolute difference (MAD), intraclass correlation coefficient (ICC [2, 1]), and Pearson correlation r.

RESULTS

All ST parameters showed excellent intra-rater reliability (ICCs [2, 1] ≥ 0.95), and good to excellent inter-rater reliability (ICC [2, 1] ≥ 0.79). Compared with XR measurements, US parameters showed a good to strong correlation with r ranging from 0.67 to 0.97, while ST parameters showed a weak to good correlation with r ranging from 0.28 to 0.73. Two ST parameters (scapular angle and coronal balance) combined with US measurements improve accuracy in assessing the right thoracic curve and axial vertebral rotation.

CONCLUSION

Combining US and ST parameters had merit in providing a more accurate assessment than using the US or ST only to estimate the Cobb angle and axial vertebral rotation on right thoracic curves.