Predicting cortical-thalamic functional connectivity using functional near-infrared spectroscopy and graph convolutional networks

• Published in: Scientific reports Vol. 14; no. 1; pp. 29794 - 17
• Main Authors: Tang, Lingkai, Kebaya, Lilian M. N., Vahidi, Homa, Meyerink, Paige, de Ribaupierre, Sandrine, Bhattacharya, Soume, St. Lawrence, Keith, Duerden, Emma G.
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 30.11.2024; Nature Publishing Group; Nature Portfolio
• Subjects: 631/114/116/1925; 631/114/1305; 692/700/1421/65; Adult; Brain Injuries - diagnostic imaging; Brain Injuries - physiopathology; Brain injury; Brain mapping; Cerebral Cortex - diagnostic imaging; Cerebral Cortex - physiology; Female; Functional connectivity; Functional magnetic resonance imaging; Functional near-infrared spectroscopy; Graph convolutional network; Humanities and Social Sciences; Humans; Infant, Newborn; Infrared spectroscopy; Machine Learning; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; multidisciplinary; Neonates; Neural networks; Neural Networks, Computer; Neuroimaging; Prediction models; Science; Science (multidisciplinary); Spectroscopy, Near-Infrared - methods; Spectrum analysis; Thalamus; Thalamus - diagnostic imaging; Thalamus - physiology; Young Adult; Functional connectivity; Functional magnetic resonance imaging; Graph convolutional network; Functional near-infrared spectroscopy; Machine learning
• DOI: 10.1038/s41598-024-79390-3
• ISSN: 2045-2322

Functional near-infrared spectroscopy (fNIRS) measures cortical hemodynamic changes, yet it cannot collect this information from subcortical structures, such as the thalamus, which is involved in several key functional networks. To address this drawback, we propose a machine-learning-based approach to predict cortical-thalamic functional connectivity using cortical fNIRS data. We applied graph convolutional networks (GCN) to two datasets obtained from healthy adults and neonates with early brain injuries, respectively. Each dataset contained fNIRS connectivity data as input to the predictive models, while the connectivity from functional magnetic resonance imaging (fMRI) served as training targets. GCN models performed better compared to conventional methods, such as support vector machine and feedforward fully connected artificial neural networks, on both identifying the connections as binary classification tasks, and regressing the quantified strengths of connections. We also propose the addition of inter-subject connections into the GCN kernels could improve performance and that GCN models are resilient to noise in fNIRS data. Our results show it is feasible to identify subcortical activity from cortical fNIRS recordings. The findings can potentially extend the use of fNIRS in clinical settings for brain monitoring in critically ill patients.