Whole process of standardization of diffusion-weighted imaging: Phantom validation and clinical application according to the QIBA profile

Abstract

Abstract Objectives: To use the apparent diffusion coefficient (ADC) as reliable biomarkers, validation of MRI equipment performance and acquisition protocols should be performed prior to application in patients. This study aims to perform validation for various MRI equipment and clinical protocols for diffusion weighted imaging (DWI) using commercially available phantom, and the validated protocols were confirmed in patients’ images. Materials and Methods: The MRI equipment performance of four different scanners and clinical brain protocols were validated using a Quantitative Imaging Biomarker Alliance (QIBA) diffusion phantom and cloud-based analysis tool. We evaluated performance metrics regarding accuracy and repeatability of ADC using QIBA profile. The validated clinical protocols were applied to 17 patients, and image quality and repeatability of ADC were assessed. Results: The MRI equipment performance of all four MRI scanners demonstrated high accuracy in ADC measurement (ADC bias, -2.3% to -0.4%), excellent linear correlation to the reference ADC value (slope, 0.9–1.0; R2, 0.999–1.000), and high short-term repeatability [within-subject-coefficient-of-variation (wCV), 0%–0.3%]. The clinical protocols were also validated by fulfilling the QIBA claims with high accuracy of ADC measurement (ADC bias, -3.1-0.7%) and robust repeatability (wCV, 0%–0.1%). DWI acquired using the validated clinical brain protocols showed ideal image quality (mean score≥2.9) and good repeatability (wCV, 1.8–2.2 for white matter, 2.0–2.1 for cerebrospinal fluid). Conclusion: The whole process of standardization of DWI demonstrated the robustness of DWI with high accuracy and repeatability across diverse MRI equipment and clinically optimized protocols, which is in accordance with the QIBA claims.