Evaluation of Whole Brain Intravoxel Incoherent Motion (IVIM) Imaging

Abstract

Intravoxel Incoherent Motion (IVIM) imaging provides non-invasive perfusion measurements, eliminating the need for contrast agents. This work explores the feasibility of IVIM imaging in whole brain perfusion studies, where an isotropic 1 mm voxel is widely accepted as a standard. This study follows the validity of a time-limited, precise, segmentation-ready whole-brain IVIM protocol suitable for clinical reality. To assess the influence of SNR on the estimation of S₀, f, D*, and D IVIM parameters, a series of measurements and simulations were performed in MATLAB for the following three estimation techniques: segmented grid search, segmented curve fitting, and one-step curve fitting, utilizing known "ground truth" and noised data. Scanner-specific SNR was estimated based on a healthy subject IVIM MRI study in a 3T scanner. Measurements were conducted for 25.6 × 25.6 × 14.4 cm FOV with a 256 × 256 in-plane resolution and 72 slices, resulting in 1 × 1 × 2 mm voxel size. Simulations were performed for 36 SNR levels around the measured SNR value. For a single voxel grid, the search algorithm mean relative error Ŝ₀, f^, D^*, and D^ of at the expected SNR level were 5.00%, 81.91%, 76.31%, and 18.34%, respectively. Analysis has shown that high-resolution IVIM imaging is possible, although there is significant variation in both accuracy and precision, depending on SNR and the chosen estimation method.

Keywords: DWI; IVIM; MRI; SNR; brain; perfusion.

Figure 1

A bi-exponential IVIM model illustration. The orange curve represents the prediction from over-threshold b-values. The blue curve represents the actual data. The difference between the prediction from high b-values and the actual data is attributed to the influence of blood circulation and is referred to as pseudo-diffusion [8]. It is seen that b-values where scanning is sensitive for fast components lie in the range of approx. 0–250.

Figure 2

(a) Histogram of SNR (left). (b) Heatmap of SNR shown as axial slices 20 mm from each other (right).

Figure 3

Gray matter coverage on DWI scan. The yellow shape represents the coverage of the gray matter map with a probability > 80% overlayed on $b_0$ image, regridded to IVIM DWI resolution.

Figure 4

RMSE of parameter fitting for three methods of single voxel estimation.

Figure 5

RMSE of parameter fitting for three methods. Estimation was made for clustered voxels and each estimation for the signal was averaged from eight voxels.

Figure 6

RMSE of parameter fitting for three methods. Estimation made for clustered voxels, each estimation for signal averaged from 27 voxels.

Figure 7

RMSE of parameter fitting for three methods. Estimation was made for clustered voxels and each estimation for the signal was averaged from 64 voxels.

Figure 8

Brodmann areas atlas overlayed on DWI scans. Red and green shape represents the coverage of the Brodmann atlas template, regridded to IVIM DWI resolution.

Figure 9

Comparison of (a) diffusion, (b) pseudo-diffusion parameter estimation in mm²/s, and (c) blood fraction parameter estimation; calculated on the signal from Brodmann areas.

Figure 10

Histogram of estimate value (x-axis) on white and grey matter (sum of >70% TPM maps) and heatmap of voxels using grid search calculation method.: (a) $f$ and (b) $D^{*}$.