Friday Lunch Seminar (English)
October 27, 2023
12:15 〜 13:00

Sing-up for participation  by noon, October 26
from here
You will be notified of participation details by e-mail on October 26.

Talk Title: Fast, faster, fastest in fMRI

Shota Hodono
Postdoctoral Research Fellow
The University of Queensland

Host PI: Hiromasa Takemura

Abstract:
In this talk, I will present our work investigating the temporal sensitivity of Blood Oxygen Level Dependent (BOLD) based fMRI, the temporal specificity of Diffusion fMRI (DfMRI), and the challenges associated with Direct Imaging of Neuronal Activity (DIANA) in humans.

Gradient recalled echo (GRE) BOLD contrast allows inference of neuronal activity from associated hemodynamic changes. Therefore, the spatiotemporal sensitivity is influenced by the local vasculature architecture. The spatial specificity has been well studied, however the temporal aspects has been less explored. We have studied the frequency dependence of the BOLD signal to stimulus-driven hemodynamic responses in primally motor and somatosensory cortices (M1 and S1) using cortical depth as a proxy for the vascular hierarchy.

DfMRI has been proposed as a mean to detect more neuronally specific signal changes, possibly associated with activity induced cell-swelling. However, the field has struggled to reproduce key findings presented in the original paper. Commonly, attempts to reproduce these findings resulted in signals indistinguishable from BOLD. Using extremely short readouts, we were able to minimize BOLD contamination in DfMRI, revealing signal preceding the BOLD response by 1.6 seconds.

The recently proposed DIANA technique claims to offer neuronally specific fMRI with millisecond temporal resolution. We have attempted DIANA in humans and rats, but failed to observe reliable normally driven signals. I will conclude my talks by sharing our initial experiences about DIANA in humans, its challenges, and related technological developments.

“