Modeling the Hemodynamic Response Function in fMRI: Efficiency, Bias and Mis-modeling

By Martin A. Lindquist, Ji Meng Loh, Lauren Y. Atlas, Tor D. Wager. NeuroImage, 2008

552 citations recorded by the source corpus.

Most brain research to date has focused on studying the amplitude of evoked fMRI responses, though there has recently been an increased interest in measuring onset, peak latency and duration of the responses as well. A number of modeling procedures provide measures of the latency and duration of fMRI responses. In this work we compare several techniques that vary in their assumptions, model complexity, and interpretation. For each method, we introduce methods for estimating amplitude, peak latency, and duration and for performing inference in a multi-subject fMRI setting. We then assess the techniques’ relative sensitivity and their propensity for mis-attributing task effects on one parameter (e.g., duration) to another (e.g., amplitude). Finally, we introduce methods for quantifying model misspecification and assessing bias and power-loss related to the choice of model. Overall, the results show that it is surprisingly difficult to accurately recover true task-evoked changes in BOLD signal and that there are substantial differences among models in terms of power, bias and parameter confusability. Because virtually all fMRI studies in cognitive and affective neuroscience employ these models, the results bear on the interpretation of hemodynamic response estimates across a wide variety of psychological and neuroscientific studies.

Subfields: Cognitive Neuroscience, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/08d63c6bbfaef087d9945218da5c99831dca6752

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Building a Science of Individual Differences from fMRI. (2016, methodology): “Some solutions have been proposed for this joint comparison, yet none is widely accepted yet [51,55].”
Interpretable LSTMs For Whole-Brain Neuroimaging Analyses (2018, result): “This finding is generally in line with the temporal evolution of the hemodynamic response function (HRF; [53]) measured by the fMRI (the HRF is known to peak 5-10 seconds after to the onset of the underlying neuronal activity).”
The Statistical Analysis of fMRI Data. (2008, background): “This is an extremely important problem that has already started to become a major area of research.”
Analyzing Neuroimaging Data Through Recurrent Deep Learning Models (2018, unknown): “To mimic the blood-oxygen-level dependent (BOLD) response measured by the fMRI, each predictor is first convolved with a hemodynamic response function (HRF; Lindquist et al., 2009), before fitting the β-coefficients to the data.”
Accurate localization of cortical and subcortical sources of M/EEG signals by a convolutional neural network with a realistic head conductivity model: Validation with M/EEG simulation, evoked potentials, and invasive recordings (2024, background): “In contrast, other non-invasive measurements, like near-infrared spectroscopy (NIRS) and functional magnetic resonance imaging (fMRI), provide a low temporal resolution on the order of seconds due to their reliance on hemodynamic responses (Lindquist et al., 2009).”
Cognitive interference elicits developmental sex differences in inhibitory control (2025, unknown): “…of time resolved methods such as magnetoencephalography (MEG) may lead to unique insights, as these tools can often distinguish elements of interference control from behavioral outputs, which may lead to a finer differentiation of the subcomponents of inhibitory control (Lindquist et al., 2009).”