Direct Quantification of Intervertebral Disc Water Content Using MRI

Water content is a key parameter for simulating tissue swelling and nutrient diffusion. Accurately measuring water content throughout the intervertebral disc (NP = nucleus pulposus; AF = annulus fibrosus) is important for developing patient-specific models. Water content is measured using destructive techniques, Quantitative MRI has been used to estimate water content and detect early degeneration, but it is dependent on scan parameters, concentration of free water molecules, and fiber architecture.To directly measure disc-tissue water content using quantitative MRI and compare MRI-based measurements with biochemical assays, and to quantify changes in disc geometry due to compression.Basic science, controlled.Twenty bone-disc-bone motion segments from skeletally mature bovines.7T/3D fast low angle shot (FLASH) pulse sequence and a T2 rapid imaging with refocused echoes (RARE) sequence.Disc volumes, NP and AF volumetric water content, and T2 relaxation times were measured through MRI; NP and AF tissue gravimetric water content, mass density, and glycosaminoglycan content were measured through a biochemical assay.Correlations between MRI-based measurement and biochemical composition were evaluated using Pearson's linear regression.Mechanical dehydration resulted in a decrease in disc volume by up to 20% and a decrease in disc height by up to 35%. Direct water content measurements for the NP was achieved by normalizing MRI-based spin density by NP mass density (1.10 ± 0.03 g/cm3 ). However, the same approach underestimated water content in the AF by ~10%, which may be due to a higher concentration of collagen fibers and bound water molecules.Spin density or spin density normalized by mass density to estimate NP and AF water content was more accurate than correlations between water content and relaxation times. Mechanical dehydration decreased disc volume and disc height, and increased maximum cross-sectional area.TECHNICAL EFFICACY STAGE: J. Magn. Reson. Imaging 2020;52:1152-1162.