Detection of Discogenic Low Back Pain using pH-Level Dependent MRI.

摘要

Discogenic low back pain (LBP) is a major medical condition in the intervertebral disc (IVD) that has wide prevalence and significant economic and social burden. The current diagnostic `gold standard', discography, is an invasive procedure and uses ionizing radiation. Furthermore, discography has high false positive rate and sometimes deteriorate the pain. Currently there are no non-invasive imaging methods that target disease specific biomarkers for detecting painful IVDs. Previous studies on pathogenesis have linked acidic IVD environment with discogenic LBP. Therefore there is a potential need for an imaging based technique that measures IVD pH level in vivo, and using this technique for detecting painful IVDs.;The purpose of this work was to develop an in vivo pH level-dependent magnetic resonance imaging (MRI) method for detecting discogenic LBP, without using exogenous contrast agents. This method uses the ratio of two MRI measurements, R1rho dispersion between two different spin-lock amplitudes and -OH (hydroxyl) chemical exchange saturation transfer (CEST).;The first aim of this study was to achieve reliable in vivo CEST and spin-lock imaging at 3.0T. CEST and spin-lock imaging are susceptible to several effects that prohibit their accurate measurements on a 3.0 Tesla MR scanner. First, the frequency-drift phenomenon during CEST study was investigated and corrected. Second, motion artifacts caused by bowel movement was investigated, and a technique that suppresses these artifacts in both CEST and spin-lock imaging using reduced-field-of-view (rFOV) turbo spin-echo (TSE) was developed. This technique was applied on healthy volunteers and was compared with conventional TSE.;The second aim of this study was to develop and validate the ratio of R1rho dispersion and -OH CEST (RROC) for IVD pH imaging. A novel acquisition scheme for in vivo R1rho measurement at low spin-lock amplitude was developed to correct for magnetic field inhomogeneities. Then the use of RROC for pH-level dependent imaging was developed and verified by numerical simulations, and by phantom and animal studies. Patients with discogenic LBP and scheduled for discography were examined with this RROC method on a 3.0 Tesla scanner. RROC measurements were compared with discography outcomes using statistical analysis to test its pain detection power.