Decoding letter position in word reading.

Area of Science:

• Neuroscience
• Cognitive Psychology
• Computational Linguistics

Background:

• Visual word recognition involves transforming letter sequences into unified word forms.
• Neuropsychological evidence suggests letter position processing can be selectively impaired, leading to peripheral dyslexia.
• The precise neural mechanisms for encoding relative letter positions remain under investigation.

Purpose of the Study:

• To investigate the neural basis of letter position processing during visual word recognition.
• To identify brain regions involved in distinguishing the order of letters within a word.
• To explore the role of the left intraparietal sulcus (IPS) in this process.

Main Methods:

• Utilized functional magnetic resonance imaging (fMRI) to measure brain activity during reading.
• Applied supervised machine learning techniques to classify letter position based on fMRI activation patterns.
• Conducted functional connectivity analysis to examine interactions between brain regions.

Main Results:

• Activity patterns in the left IPS accurately classified letter position (80% accuracy).
• Consistent classification performance was observed across different letters of interest within the same voxels.
• Functional connectivity revealed co-activation between the left IPS and the Visual Word Form Area during reading.

Conclusions:

• The left IPS plays a significant role in processing relative letter positions within the reading network.
• Findings align with clinical observations of letter position dyslexia associated with left occipito-parietal lesions.
• This study provides novel evidence for the neural representation of letter order in reading.