# A comparative analysis of the sensitivity and BOLD contamination of the VASO response at 3 Tesla: ME-DEPICTING vs. ME-EPI readouts

## TL;DR

The study compares two MRI techniques for measuring blood volume changes in the brain, finding that signal corrections are crucial to avoid overestimation.

## Contribution

The paper introduces a novel comparison of ME-DEPICTING and ME-EPI for VASO signal correction and sensitivity at 3 Tesla.

## Key findings

- Uncorrected VASO data from the first echo showed significant differences in sensitivity between the two readouts.
- ME extrapolation corrected VASO data showed similar sensitivity for detecting cerebral blood volume changes.
- Dynamic division correction led to a near-perfect linear dependence on echo time due to intravascular BOLD contributions.

## Abstract

‘Non-BOLD fMRI’ data acquired at non-zero echo time (TE) suffer from contamination by the Blood Oxygenation Level Dependent (BOLD) signal due to the unavoidable signal decay caused by transverse relaxation. This contamination further reduces their already low inherent functional sensitivities and makes their correction essential. The Slice-Saturation Slab-Inversion Vascular Space Occupancy (SS-SI–VASO), for instance, cancels out BOLD contributions from VASO data, reflecting cerebral blood volume (CBV) changes, via a dynamic division approach. Alternatively, multi-echo (ME) data provide the possibility of extrapolating toTE=0. Acquisitions at very shortTEwould minimize the need for such corrections. The center-out EPI variant (‘DEPICTING’) is one such readout which allows for shortTE. The ME 2D DEPICTING was compared here against a traditional ME 2D EPI for its sensitivity to functional changes in the VASO signal. The two BOLD-correction schemes were also evaluated. Clear differences in functional sensitivity were observed for the uncorrected VASO data obtained from the first echo,TE1, of the two readouts. VASO data corrected by ME extrapolation were, however, found to be almost identical in their sensitivity for detecting CBV changes for both readouts. An excessively high increase in VASO signal sensitivity observed with the dynamic division correction for both readouts revealed a near-perfect linear dependence onTEof VASO signal changes. This could be attributed to the substantial intravascular BOLD contributions at 3 T. In the present data, extravascularΔR2*fraction was found to be around ~50–60%. ME extrapolation is, hence, recommended to avoid overestimation of functional CBV changes at commonly used TEs.