Translating the hemodynamic response: why focused interdisciplinary integration should matter for the future of functional neuroimaging

Abstract

The amount of information acquired with functional neuroimaging techniques, particularly fNIRS and fMRI, is rapidly growing and has enormous potential for studying human brain functioning. Therefore, many scientists focus on solving computational neuroimaging and Big Data issues to advance the discipline. However, the main obstacle-the accurate translation of the hemodynamic response (HR) by the investigation of a physiological phenomenon called neurovascular coupling-is still not fully overcome and, more importantly, often overlooked in this context. This article provides a brief and critical overview of significant findings from cellular biology and in vivo brain physiology with a focus on advancing existing HR modelling paradigms. A brief historical timeline of these disciplines of neuroscience is presented for readers to grasp the concept better, and some possible solutions for further scientific discussion are provided.

Keywords: Brain; Cerebrovascular regulation; Computational modelling; Functional magnetic resonance imaging; Functional near-infrared spectroscopy; Healthcare; Hemodynamic response; Neuroscience; Neurovascular coupling.

Figure 1. Examples of canonical hemodynamic response (A), and hemodynamic response function (B).

A neural activity from 0 to 5 s (grey bar) causes neurometabolic and later neurovascular coupling, which can be seen as a delay of response (around 2 s). Box in hemodynamic response (A) indicates (a) small inflow of Δ(HbO₂), when the total blood volume is still relatively unchanged (due to increased cerebral blood flow), and later (b) Δ(HbO₂) increases rapidly due to functional hyperemia caused by vasodilatation. The small increase of Δ(Hb) occurs due to insufficient washout when the cellular oxygen demand exceeds current supply in a tissue. The canonical example of a hemodynamic response is based on measuring the composition of cerebral blood volume via chromophore concentration changes (oxy-Hb and deoxy-Hb). fNIRS studies can directly measure both oxy-Hb and deoxy-Hb), (Venclove, Daktariunas & Ruksenas, 2015). In contrast, the canonical hemodynamic response function (HRF) from the Blood Oxygenation Level Dependent (BOLD) method represents the magnetic field change in response to the Δ(Hb) curve (B) and is relative to the baseline.

Figure 2. Schematic representation of the Neurovascular Unit (NVU).

Figure 3. The conceptual biophysical scheme of biological signal transduction path in the Neurovascular Unit (NVU).

(A) Neurometabolic coupling (NMC); (B) Neurovascular coupling (NVC). Both neurones (neurotransmitter release) and astrocytes (glucose and oxygen consumption) respond to increased extracellular glutamate, and intracellular calcium to transmit direct and indirect vasoactive signals for the appropriate blood delivery and distribution in the electrically active brain area.

Figure 4. A timeline of magnetic resonance imaging (MRI), functional near-infrared spectroscopy (fNIRS), and cellular and molecular neuroscience (CMN) milestones.