UPmémoires

English

Objective : This study focused on evaluating the validity and test characteristics of TOF-camera in patients with idiopathic scoliosis (IS). This study aimed 1) to evaluate the characteristics of GTA measured with TOF-camera as a tool for describing AIS; 2) to compare these results with those obtained with the scoliometer.

Summary of Background Data : Definition of scoliosis is based on the x-rays with a lateral spinal curvature in the frontal plane. The overall prevalence is 3% in the population between 10 and 18 years. The close monitoring of IS population during the growing years is necessary and requires repetitive x-rays, which could lead to long-term health problems. Finding radiation free tools for screening and monitoring IS is therefore a major challenge. The most common test used for screening is based on the evaluation of the angle of trunk rotation (ATR) with a scoliometer during forward bending test, but it is not generally used to assess curvature evolution.

Patients and methods : Patients suspected of IS scheduled for x-ray were screened. To evaluate the validity of the global trunk asymmetry (GTA) provided by a TOF-camera, a comparison (Pearson correlation) was performed with the major Cobb angle (CA) obtained using 2D EOS® radiography. GTA and ATR measured with a scoliometer were evaluated. One hundred and eight patients aged from 5 to 24 (mean 13,9 years) were seen during spine specialist consultation. A photograph with the TOF-camera was performed for the 108 patients and a measure of ATR with a scoliometer during Adam's forward bending test, was performed for 93 of them. The AUC statistics from ROC curves were used to describe the discriminative ability of the GTA and ATR parameters with the IS diagnosis defined with a Cobb angle>10° (X-Ray). Test characteristics (sensibility and specificity) were reported for the best threshold, identified using the Youden method, for a CA>10° as well as a CA>20° considered as a treatment cutoff. AUC of GTA and ATR were compared using a Bootstrap test for two correlated ROC curves methods.

Results : Considering scoliosis diagnosis (CA≥10°), AUC was 0,88 (0,81-0,96) for GTA (threshold=7°, Se=81%, Sp=83%) and 0,92 (0,85-0,99) for ATR. The AUC of GTA and ATR were not significantly different (p=0.399). Considering a CA≥20°, as a cut-off for orthopaedic treatment, AUC was 0,88 (95%CI=0,82-0,95) for GTA (threshold=7,87, Se =87%, Sp=75%).

Conclusions : The use of a TOF-camera with thresholds of 7 and 7,87, seem to offer interesting screening characteristics of IS. Based on these results, this fast, light, and non-invasive system is a promising tool for IS evaluation. It appears obvious that it needs further investigations such as solid screening characteristics, inter and intra-individual reliability and validity, to be used for monitoring the evolution of IS in growing patients instead of repetitive radiographs.

Keywords : TOF-camera, spine surface topography, idiopathic scoliosis, validity, test characteristics, global trunk asymmetry.