Machine Learning for Brain Images Classification of Two Language Speakers.

Area of Science:

• Neuroscience
• Artificial Intelligence
• Medical Imaging

Background:

• Brain structure, particularly white matter, shows differences in individuals learning a second language.
• Machine learning (ML) offers potential for analyzing complex biological data like brain images.

Purpose of the Study:

• To classify Magnetic Resonance Images (MRIs) of bilingual and monolingual individuals using ML.
• To investigate ML's capability in automatically detecting differences in brain structure related to bilingualism.

Main Methods:

• Testing various artificial neural networks with a single hidden layer, optimizing to two neurons.
• Utilizing nine hundred input features for the ML classifier.
• Employing supervised training for the neural network models.

Main Results:

• The ML classifier achieved a high percentage of effectiveness in distinguishing between bilingual and monolingual subjects.
• The study demonstrated the feasibility of automated analysis of brain MRIs for detecting bilingualism.

Conclusions:

• ML provides a viable, quantitative alternative to manual analysis methods like Tract-Based Spatial Statistics for studying brain structure.
• This research contributes to neuroscience by enabling automatic detection of bilingualism from MRIs, supporting existing findings.