Comprehensive Summary
Schizophrenia is a neuropsychiatric disorder that is characterized by extensive dysfunction in the prefrontal-limbic system and dysregulation of brain network connectivity. This study examined the use of functional near-infrared spectroscopy (fNIRS) as a tool to monitor neural activity in schizophrenia patients, and its integration into multimodal frameworks for diagnosis and treatment. fNIRS is a non-invasive imaging technique that measures changes in oxygenated (HbO) and deoxygenated (HbR) cerebral hemoglobin concentrations by taking absorption spectra of HbO and HbR, inferring local hemodynamic responses. fNIRS offers key advantages like high portability, resistance to motion interference, and real-time monitoring ability, which makes it more fitting for schizophrenia patients who cannot tolerate fMRI or electrocardiogram (EEG) constraints during psychotic states, especially due to motor restlessness. fNIRS can also capture abnormalities in resting-state brain networks, higher cognitive function deficits, and evaluate neuroplasticity, providing a different perspective to reveal neural mechanisms of schizophrenia; when combined with neuromodulation techniques, fNIRS can improve negative symptoms and offer individualized diagnosis and treatment options. Since fNIRS emits near-infrared light (650-950 nm) to penetrate through the the scalp and skull, fNIRS mainly only detects changes in superficial cortical regions like the prefrontal cortex, making its spatial resolution and detection of whole-brain activity more limited than fMRI. Signal noise and methodological heterogeneity are also some limitations. However, fNIRS’s flexibility to head movement and its comprehensive and cost-effective qualities make it a practical and valuable complement to traditional neuroimaging techniques. Future research should focus on the integration of AI with multimodal imaging techniques, standardization of task design, and developing adaptive noise suppression algorithms.
Outcomes and Implications
fNIRS offers a real-time way to monitor prefrontal and temporal function in schizophrenia by offering hemodynamic data that is tolerant to motions such as head movement, which is particularly valuable in individuals experiencing psychosis. fNIRS neurofeedback treatment has been shown to reduce treatment-resistant auditory verbal hallucinations, especially because HbO and HbR concentration changes often indicate changes in symptoms, so fNIRS data can serve as an early biomarker of responses to different interventions like antipsychotics, psychotherapy, neurofeedback, and other tasks. Machine learning (ML) has also demonstrated high precision in fNIRS signal classification, with deep neural networks (DNNs) being able to distinguish first-episode schizophrenia from healthy individuals with an AUC of 0.89, and dynamic functional connectivity (DFC) achieving up to 82% accuracy in differentiating schizophrenia from bipolar disorder, suggesting that fNIRS-based data can aid in refining differential diagnosis. Some limitations of fNIRS like shallow cortical depth, heterogeneity of methods, and signal noise, means that fNIRS could be integrated as an aiding tool in multimodal imaging with EEG or fMRI, improving both spatial and temporal resolution and allowing for continuous tracking of treatment outcomes. These advances posit fNIRS as a clinically feasible tool for personalized treatment monitoring, neuromodulation guidance, and validation of ML biomarkers that could aid in early treatment diagnoses for schizophrenia.