Objective: This study aimed to validate the accuracy of a scoliosis assessment method using a commercially available 3D depth camera (Azure Kinect). By introducing a correction coefficient to minimize systematic errors, the research sought to enhance both practicality and accuracy in daily scoliosis evaluation.
Background: With the increasing prevalence of scoliosis due to prolonged use of digital devices, there is growing demand for accessible, cost-effective, and non-invasive tools for spinal health monitoring. Traditional methods, such as the scoliometer, are limited to clinical settings and lack suitability for frequent self-assessment. This study highlights the practical contribution of applying a consumer-grade depth camera to scoliosis assessment, demonstrating its potential for widespread daily use.
Method: A 3D depth camera was employed to evaluate the degree of spinal curvature using the Adams forward bend test. Sixteen participants were recruited, and assessments were conducted in two postures: a normal forward-bending position (N = 16) and an induced scoliosis posture (N = 10). Each posture was measured using both the 3D depth camera and a conventional scoliometer. The consistency between the two measurement methods was analyzed using the Intraclass Correlation Coefficient (ICC).
Results: Strong agreement was observed between the depth camera and scoliometer results (ICC = 0.914, p < 0.05). After applying the correction coefficient, accuracy improved from 78% to 91%, with absolute agreement reaching ICC = 0.996. These findings confirm both the reliability and practical efficiency of the proposed method.
Conclusion & Application: The scoliosis assessment method using a 3D depth camera provides a highly accessible, non-invasive, and accurate alternative to traditional clinical tools. Its practical advantages-low cost, ease of use, and strong accuracy-highlight its value as a digital healthcare solution for routine and preventive spinal health management.