Identification of significant quantitative MRI features for lumbar intervertebral disc degeneration through comprehensive image analysis

Motivation: Visual analysis has been the mainstay in evaluating intervertebral disc (IVD) degeneration. A comprehensive radiomic analysis of quantitative MRI features may improve distinguishing IVDs by severity of degeneration. Goal(s): To identify significant quantitative MRI features for lumbar IVD degeneration through comprehensive image analysis Identifying effective MRI quantitative imaging features for evaluating intervertebral disc degeneration Approach: Histogram and texture features of ADC, T2*, and T1ρ were extracted, and features distinguishing severe IVD degeneration were identified by decision trees. Results: Four texture and one histogram features were identified, as significant in distinguishing the severity of IVD degeneration. Impact: This study highlights that incorporating texture features of quantitative MRI indices (ADC, T2*, and T1ρ) can improve determination of the severity of IVD degeneration.