Comparison of different analytical methods used for analyzing SPECT scans of patients with unilateral condylar hyperactivity

Abstract

This study aimed to compare different analytical methods and identify the optimal analysis method to distinguish patients with unilateral condylar hyperactivity (UCH) from those with inactive condyles. Single-photon emission computed tomography (SPECT) scans of patients with progressive and nonprogressive mandibular asymmetry (each group, n=26), were analyzed using the region of interest technique. Sensitivity, specificity and receiver operating characteristic (ROC) curves were calculated for the different analytic methods. The ROC curve illustrates that UCH can be diagnosed significantly better by determining the percentile bone activity in both condyles. The area under the curve (AUC) of the percentile comparison between the affected and contralateral condyles was 0.93+/-0.04, that for the condyle/clivus ratio was 0.75+/-0.07 and for the condyle/cervical spine (CS) ratio 0.57+/-0.08. Sensitivity for the condyle/clivus ratio was 65% and specificity 61%. Sensitivity for the condyle/CS ratio was 85% and specificity 31%. For the percentile difference of the condyles, sensitivity and specificity were 88%. For UCH patients, direct comparison of bone activity between the affected and contralateral condyle in SPECT scans is the analysis method of choice. Comparison of condylar bone activity to reference bone activity does not have additional value in the diagnosis of UCH.