Neural networks supporting audiovisual integration for speech: A large-scale lesion study

Area of Science:

• Neuroscience
• Cognitive Science
• Audiology

Background:

• Auditory and visual speech information are typically integrated for enhanced perception.
• Mismatched audiovisual cues can lead to illusory percepts, such as the McGurk-MacDonald effect.
• Previous research implicated the posterior superior temporal sulcus (STS), auditory cortex, and inferior frontal gyrus in audiovisual speech integration.

Purpose of the Study:

• To investigate the causal involvement of brain regions in audiovisual (AV) speech integration using a large-scale lesion-based study.
• To determine if different measures of AV speech integration rely on the same neural mechanisms.
• To identify brain areas critical for AV speech enhancement and illusory fusion.

Main Methods:

• Conducted the first large-scale lesion-based study (N=100) on AV speech integration.
• Analyzed behavioral performance and lesion maps for AV enhancement and illusory fusion.
• Correlated lesion locations with specific types of AV speech processing failures (auditory vs. visual capture).

Main Results:

• Classic metrics for AV speech integration may not capture the same underlying processes.
• Lesions in superior temporal auditory, lateral occipital visual, and multisensory STS zones significantly disrupt AV speech integration.
• The location of brain lesions predicts whether auditory or visual information dominates when AV speech integration fails.

Conclusions:

• AV speech processing is supported by both unimodal auditory/visual cortices and multimodal regions like the STS.
• Motor-related frontal regions do not appear to be essential for AV speech integration.
• Understanding lesion effects provides insight into the neural basis of multisensory speech perception.