Multivariate analysis of cartilage degradation using the support vector machine algorithm

Abstract

An important limitation in MRI studies of early osteoarthritis is that measured MRI parameters exhibit substantial overlap between different degrees of cartilage degradation. We investigated whether multivariate support vector machine analysis would permit improved tissue characterization. Bovine nasal cartilage samples were subjected to pathomimetic degradation and their T(1), T(2), magnetization transfer rate (k(m) ), and apparent diffusion coefficient (ADC) were measured. Support vector machine analysis performed using certain parameter combinations exhibited particularly favorable classification properties. The areas under the receiver operating characteristic (ROC) curve for detection of extensive and mild degradation were 1.00 and 0.94, respectively, using the set (T(1), k(m), ADC), compared with 0.97 and 0.60 using T(1), the best univariate classifier. Furthermore, a degradation probability for each sample, derived from the support vector machine formalism using the parameter set (T(1), k(m), ADC), demonstrated much stronger correlations (r(2) = 0.79-0.88) with direct measurements of tissue biochemical components than did even the best-performing individual MRI parameter, T(1) (r(2) = 0.53-0.64). These results, combined with our previous investigation of Gaussian cluster-based tissue discrimination, indicate that the combinations (T(1), k(m)) and (T(1), k(m), ADC) may emerge as particularly useful for characterization of early cartilage degradation.