A nomogram for classification of temporomandibular joint disk perforation based on magnetic resonance imaging

Abstract

The aim of this study was to analyze risk factors and establish a prediction model for temporomandibular joint (TMJ) disk perforation by constructing a nomogram. The study included a total of 282 joints in 274 patients. All patients underwent open TMJ surgery after obtaining magnetic resonance imaging (MRI), from 2005 to 2015. The presence or absence of disk perforation was confirmed during the operation. Patients were classified into 2 groups: perforation and nonperforation groups. We investigated demographic data and the characteristics of the disk, joint space, and bone on MRI. A logistic regression analysis was performed to analyze risk factors. A nomogram was constructed and validated internally and externally. Risk factors for disk perforation were increased age, disk shape (eyeglass or amorphous), low bone marrow signal, abnormal joint space, and 2 or more bony changes in the condyle and fossa. The area under the receiver operating characteristic curve of the nomogram was 0.908 (95% confidence interval [CI] 0.869-0.946) in the internal validation and 0.889 (95% CI 0.804-0.973) in the external validation with good suitability. We were able to predict the probability of disk perforation with analyzed risk factors and constructed a nomogram, which may be helpful in proper diagnosis and treatment.