Repeatability of diffusion‐weighted imaging in rectal cancer

Abstract

Serial diffusion-weighted MRI (DW-MRI) measurements of the apparent diffusion coefficient (ADC) of rectal tumors are used for rectal cancer response evaluation after neo-adjuvant treatment. In this study, we determined the repeatability of DW-MRI to distinguish therapy-related response from measurement variations. In 18 patients with rectal cancer on five consecutive days, 1.5 Tesla (T) MR imaging was performed including two identical DW-MRI sequences. The repeatability of the tumor ADC measurements and the intraobserver ADC variation were depicted in a Bland-Altman plot. The repeatability coefficient was calculated as the range of ADC values of two identical DWI measurements for 95% of subjects. It was expressed as percentage of the mean ADC value. Three females and 15 males were included. The mean tumor ADC value was 1.15 × 10(-3) mm(2)/s (SD 0.07 × 10(-3) mm(2)). The repeatability coefficient of the ADC value was 9.8% and for the intraobserver repeatability 4.7%. In serial DW-MRI for rectal cancer treatment response evaluation, a repeatability coefficient of 9.8% has to be considered to account for measurement variations in rectal tumor ADC. These variations represent observer judgement and patient and MR spectrometer induced variations.