Related Experiment Video
Updated: May 22, 2026
05:33
How to Calculate and Validate Inter-brain Synchronization in a fNIRS Hyperscanning Study
Published on: September 8, 2021
RLS adaptive filtering for physiological interference reduction in NIRS brain activity measurement: a Monte Carlo
1School of Electrical Engineering and Automation, Harbin Institute of Technology, No. 92 West Da-zhi Street, Nangang District, Harbin, People's Republic of China. zyhit@hit.edu.cn
Physiological Measurement
|May 4, 2012
Summary
This study introduces a recursive least-squares (RLS) algorithm to reduce physiological interference in near-infrared spectroscopy (NIRS) brain measurements. The RLS algorithm demonstrates superior performance over LMS in minimizing interference and improving signal quality for functional haemodynamics.
Area of Science:
- Biomedical Engineering
- Neuroscience
- Optical Imaging
Background:
- Near-infrared spectroscopy (NIRS) enables non-invasive measurement of cerebral functional haemodynamics.
- Physiological interference significantly degrades NIRS signal quality, hindering accurate brain activity assessment.
- Effective suppression of interference is critical for reliable NIRS-based neuroimaging.
Purpose of the Study:
- To develop and evaluate an adaptive filtering algorithm for reducing physiological interference in NIRS measurements.
- To compare the performance of the recursive least-squares (RLS) algorithm against the least mean squares (LMS) algorithm for interference suppression.
- To investigate the impact of NIRS probe configuration and superficial layer thickness on algorithm performance.
Main Methods:
- Monte Carlo simulations were performed using a five-layer adult head model.
- Multidistance source-detector arrangements (short and long pairs) were employed for NIRS simulation.
- Recursive least-squares (RLS) and least mean squares (LMS) adaptive filtering algorithms were implemented for interference removal.
Main Results:
- The RLS algorithm exhibited faster convergence and a smaller mean squared error (MSE) compared to the LMS algorithm.
- RLS demonstrated greater capability in minimizing physiological interference in NIRS signals.
- Near-detector positioning and short source-detector separations (<9 mm) were found to be crucial for minimizing MSE and ensuring robustness to superficial layer variations.
Conclusions:
- The RLS algorithm offers a more effective solution for mitigating physiological interference in NIRS brain measurements.
- Optimizing NIRS probe configuration, specifically interoptode distance, is essential for enhancing signal quality.
- The findings provide valuable insights for improving NIRS instrumentation and data analysis for neuroimaging applications.
