Multi-system repeatability and reproducibility of apparent diffusion coefficient measurement using an ice-water...

Area of Science:

• Medical Imaging
• Biomarker Development
• Quantitative MRI

Background:

• Apparent diffusion coefficient (ADC) is a promising biomarker for multi-center clinical trials.
• Technical variability in ADC measurements can hinder multi-center study reliability.
• Standardized quality control procedures are essential for robust multi-center biomarker applications.

Purpose of the Study:

• To establish quantitative quality control (QC) procedures for multi-center ADC measurements.
• To evaluate and minimize technical variability in diffusion coefficient measurements.
• To ensure the reliable use of ADC as a biomarker in multi-center clinical trials.

Main Methods:

• A uniform diffusion weighted imaging (DWI) protocol was developed and distributed to 18 test sites.
• A temperature-controlled diffusion phantom was used for data acquisition across 35 MRI systems (1.5T and 3T).
• Data from ice water diffusion imaging at five b-values were centrally analyzed.

Main Results:

• Inter-scanner ADC variability was less than 2% at the bore center, with an error range of -2% to +5% from the literature value.
• Day-to-day and intra-exam repeatability were within 4.5% and 1%, respectively.
• Inter-site reproducibility was within 3% at the magnet isocenter, but off-center measurements revealed significant vendor- and system-specific spatial ADC nonuniformity bias (>10%), linked to gradient nonlinearity.

Conclusions:

• Standardization of DWI protocols significantly enhances the reproducibility of ADC measurements.
• Spatial ADC nonuniformity was identified as a critical source of error in multi-site studies.
• Implementing these QC procedures is crucial for reliable multi-center clinical trials using ADC.