Using precise word timing information improves decoding accuracy in a multiband-accelerated multimodal reading...

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Magnetic Resonance Imaging

Background:

• The blood-oxygen-level-dependent (BOLD) signal in fMRI is typically considered slow for rapid neural processes like sentence comprehension.
• Recent research suggests that very short repetition times (TRs) can yield valuable fine-grained temporal information beyond improved signal-to-noise ratio.

Purpose of the Study:

• To investigate the utility of ultra-fast TRs in functional magnetic resonance imaging (fMRI) for capturing rapid neural dynamics during language processing.
• To assess the impact of different magnetic field strengths (3T vs. 7T) and TRs on word decoding accuracy.

Main Methods:

• Utilized multiband-accelerated fMRI to acquire whole-brain data at 3-mm resolution with a TR of 500 ms.
• Collected data at both 3 Tesla (T) and 7T magnetic field strengths.
• Analyzed word decoding accuracy based on the timing of word presentation during sentence comprehension tasks.

Main Results:

• Achieved significant improvements in word decoding accuracy by leveraging word timing information from fMRI data acquired with fast TRs.
• Demonstrated superior word decoding performance at 7T compared to 3T.
• Found that word timing information extraction benefits significantly from TRs up to 1000 ms, with diminishing returns thereafter.

Conclusions:

• Ultra-fast TRs in fMRI, particularly at 7T, provide valuable temporal resolution for decoding neural representations of spoken words.
• The findings challenge the notion of BOLD signal sluggishness and highlight the potential of rapid fMRI acquisition for studying fast cognitive processes.
• Optimizing TR is crucial for maximizing temporal information extraction in fMRI studies of language.