Testing for neural responses during temporal components of trials with BOLD fMRI

Area of Science:

• Cognitive Neuroscience
• Neuroimaging
• Biophysics

Background:

• Cognitive neuroscience hypotheses often require understanding neural responses within specific temporal components of behavioral trials.
• Blood-oxygen-level-dependent functional magnetic resonance imaging (BOLD fMRI) presents challenges in precisely identifying neural activity during these temporal components due to hemodynamic transformation.
• The transformation of neural signals to fMRI signals can lead to the loss of information, making it difficult to unambiguously determine neural responses within specific time windows.

Purpose of the Study:

• To discuss the theory and methodology for addressing the challenge of inferring neural responses during specific temporal components using BOLD fMRI.
• To propose a method that utilizes a linear-time invariant model and constraints on neural waveforms to improve the specificity of fMRI analysis.
• To demonstrate the application of this method using fMRI data from a delayed-response trial.

Main Methods:

• Adoption of a linear-time invariant (LTI) model to represent the transformation from neural signals to the fMRI BOLD signal.
• Constraining the space of possible underlying neural waveforms to improve the interpretability of the fMRI signal.
• Application and validation of the proposed method on fMRI time series data acquired during a delayed-response task.

Main Results:

• The developed method, based on LTI modeling and neural waveform constraints, shows potential for overcoming information loss in hemodynamic transformation.
• The application to delayed-response trial data illustrates the feasibility of using the method to investigate neural activity during specific temporal components.
• The study provides a framework for more precise inference of neural activity timing from fMRI data.

Conclusions:

• The proposed method offers a viable approach to test cognitive neuroscientific hypotheses concerning neural responses during specific temporal components of behavioral trials using BOLD fMRI.
• By employing an LTI model and constraining neural waveforms, researchers can better infer neural activity timing and overcome limitations of standard fMRI analysis.
• This work advances the utility of fMRI in cognitive neuroscience by enabling more specific investigations into the temporal dynamics of neural processing.