NeuroLingua: An Interpretable Machine Learning Method for Bilingual Speech Reconstruction from Stereotactic EEG Signals

Recent studies in decoding neural signals for speech-related applications have shown considerable promise for advanced brain-computer interfaces (BCIs). However, most studies have focused on speech production, while auditory speech reconstruction remains a challenging task. This paper introduces NeuroLingua, a lightweight and interpretable machine learning framework for bilingual auditory speech reconstruction from stereotactic electroencephalography (sEEG) signals. While high-frequency sEEG features are often used exclusively, we propose to integrate both low- and high-frequency neural features that complement one another, and employ an extreme gradient boosting (XGBoost) regression model paired with Shapley additive explanations (SHAP) for enhanced interpretability. To evaluate NeuroLingua, we collected and analyzed a bilingual sEEG-audio dataset from epilepsy patients undergoing intracranial monitoring. We show that the proposed framework consistently outperforms conventional single-band approaches in speech reconstruction. Furthermore, the model allows us to identify the most informative neural channels for bilingual speech reconstruction tasks. This study advances the neural speech decoding studies that support the development of next-generation BCIs for assistive communication and rehabilitation in multilingual populations. Code is publicly available (https://github.com/seegdecoding/NeuroLingua).