Related Experiment Video
Updated: Jan 14, 2026
10:07
Using Fiberless, Wearable fNIRS to Monitor Brain Activity in Real-world Cognitive Tasks
Published on: December 2, 2015
27.8K
Machine Learning Prediction of the Cognitive Responses to Transcranial Prefrontal Photobiomodulation in Individuals
Summary
Researchers developed a machine learning model to predict which individuals with bipolar disorder will benefit from transcranial photobiomodulation (tPBM) therapy. Pre-treatment brain oxygenation patterns accurately identified responders, paving the way for personalized cognitive treatments.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Psychiatry
Background:
- Bipolar disorder is associated with prefrontal cortex dysfunction and cognitive decline.
- Current treatments for bipolar disorder primarily address mood stabilization, not cognitive deficits or brain oxygenation.
- Transcranial photobiomodulation (tPBM) shows potential for improving prefrontal cortex oxygenation and cognition.
Purpose of the Study:
- To develop a predictive machine learning model for identifying individuals with bipolar disorder who respond to tPBM treatment.
- To utilize functional near-infrared spectroscopy (fNIRS) data to predict cognitive improvements after tPBM.
- To identify potential hemodynamic biomarkers for tPBM efficacy.
Main Methods:
- Support vector machine (SVM) model trained on fNIRS data (528 features) from 29 participants with bipolar disorder.
- Feature selection using Fisher-scores and regularization techniques (ridge regularization, Leave-One-Subject-Out Cross-Validation) to prevent overfitting.
- Analysis of pre-treatment hemodynamic features, including total hemoglobin concentration and oxygenated hemoglobin correlations.
Main Results:
- The predictive model achieved 85% or higher accuracy and F1-score in predicting cognitive response to tPBM using 10-20 extracted fNIRS features.
- Hemodynamic features, specifically total hemoglobin concentration statistics and oxygenated hemoglobin correlations in ventrolateral prefrontal regions, were identified as crucial predictors.
- The findings highlight the role of prefrontal cortical hemodynamics in cognitive response to tPBM.
Conclusions:
- Pre-treatment fNIRS-derived hemodynamic features can serve as reliable biomarkers for predicting tPBM treatment efficacy in bipolar disorder.
- This predictive capability can enable more personalized and effective treatment strategies for cognitive deficits in bipolar disorder.
- The study provides insights into the neurobiological mechanisms underlying tPBM-induced cognitive improvements in psychiatric populations.
