Diffusion tensor imaging of the brain in a healthy adult population: Normative values and measurement reproducibility at 3 T and 1.5 T

Abstract

Diffusion tensor imaging (DTI) is an increasingly used method for investigation of brain white matter integrity in both research and clinical applications. Familiarity with normal variation of fractional anisotropy (FA) and apparent diffusion coefficient (ADC) values and measurement reproducibility is essential when DTI measurements are interpreted in clinical patients. To establish normal values for FA and ADC in a healthy adult population at 1.5 T and 3 T MRI based on region of interest (ROI) analysis, and to study the inter- and intra-observer reproducibility of the measurements. Forty healthy volunteers (26 women, 14 men, mean age 38.3, SD 11.6 years) underwent conventional MRI and DTI of the brain, 30 with 3 T and 10 with 1.5 T clinical scanners. ROI-based measurements for FA and ADC values were performed in five different anatomic locations of each hemisphere and in three locations within the corpus callosum. Mean values for FA and ADC for each region were calculated. Inter-observer variation of ROI measurements was evaluated by comparing the results of the two observers, intra-observer variation by repeated measurement of 10 subjects by both observers. The FA values varied considerably between different regions. The highest values were found in the genu and splenium of the corpus callosum and the lowest in the corona radiata, respectively. In general, ADC values showed less variation; the highest values were found in the body of the corpus callosum and the lowest in the corona radiata. The reproducibility of both inter- and intra-observer measurements also varied regionally. The highest agreement was found for the corpus callosum and the lowest for the corona radiata and centrum semiovale. In a normal adult population FA and ADC values of the brain white matter show regional variation. The repeatability of the ROI measurements also varies regionally. This regional variability must be acknowledged when these measurements are interpreted in clinical patients.