Diffusion MRI of the Abdomen

This study will develop and validate novel DW-MRI methods with unprecedented robustness to motion, favorable image quality, and quantitative precision for abdominal imaging. Upon successful completion, these methods will have broad applications, including the assessment of cancer, fibrosis and other disease processes in various abdominal organs such as the liver, pancreas, kidneys, bowel and beyond.

The primary objective is to demonstrate precise quantitative diffusion parameter mapping that is achieved by the novel, motion-robust, low-distortion DW-MRI methods in a representative and clinically relevant application, for the assessment of liver metastatic disease. Specifically, the investigators will:

1. Compare the repeatability of DW-MRI methods by calculating the squared difference between each pair of repeated ADC measurements in lesions and healthy tissue, and will model this value including the DW-MRI method as a covariate using Generalized Estimating Equations (GEE).

Secondary objectives include optimization of the methods in healthy volunteers, assessment of image quality and distortions, as well gathering preliminary data for assessment of sensitivity and specificity for detection of lesions:

1. Optimization in healthy volunteers
2. (Aim 1) Optimized bh, M1 and M2 parameters by minimization of the mean squared error and bias of ADC quantification across the liver
3. (Aim 2) Optimized motion-corrected averaging via image quality assessment by three radiologists using a Likert scale.
4. SNR (signal-to-noise ratio) will be evaluated for each DW-MRI dataset, using an expectation-maximization method, accounting for parallel imaging, spatially varying noise, and magnitude operation.
5. To assess image distortions in DW-MRI, the cross-correlation coefficient (CCC) will be used to assess alignment between each of the DW-MRI datasets and the reference T2-weighted acquisition.
6. Each DW-MRI reconstruction will be evaluated by three radiologists using a Likert scale between 0 (worst/non-diagnostic) and 4 (best) for several criteria: motion artifacts, spatial resolution, distortions, apparent SNR, and overall image quality. The post-contrast images will serve as a guide to assess for artifacts by demonstrating the liver and the lesions.
7. Per-lesion sensitivity, specificity, and accuracy will be assessed for each DW-MRI method. McNemar's test will be used to compare the sensitivity and specificity between methods.
8. Intra-reader variability will be assessed by each reader will repeating ADC measurements after two months.
9. Inter-reader variability will be assessed in ADC measurements by comparing matching lesions, based on the recorded lesion location across readers.

Specific Aims Aim 1: Optimize a reliable, motion-robust DW-MRI of the abdomen in healthy volunteers.

Aim 2: Optimize, in healthy volunteers, a high-resolution, low-distortion, motion-robust DW-MRI of the abdomen through the synergistic combination of motion-robust DW-MRI with state-of-the-art low-distortion techniques.

Aim 3: Demonstrate excellent image quality and precise quantitative diffusion parameter mapping using the novel DW-MRI methods in a representative and clinically relevant application for the assessment of liver metastases, by evaluating in patients:

3a. Quantitative and subjective image quality metrics. 3b. Precision (test-retest repeatability) of ADC measurements in the liver (including healthy parenchyma and lesions) by novel vs. standard DW-MRI methods.