Spine and back-shape changes in scoliosis.

Thirty-five untreated patients with scoliosis were studied by Cobb radiography and back-surface photogrammetry on each of at least two clinic visits. Also, the maximum vertebral axial rotation was measured. Each pair of patient visits was classified as showing either a progression, improvement, or no change in the scoliosis using a threshold of 5 degrees change in the Cobb angle. The computer Cobb measurement was the most precise measurement, and the back-surface measurements were the least precise. The Cobb measurement of scoliosis is the most precise for detecting small changes, but can be improved by a computerized analysis of digitized radiographs.