Functional near infrared spectroscopy (fNIRS) is an emerging optical neuroimaging technology that indirectly measures neuronal activity in the cortex via neurovascular coupling. It quantifies hemoglobin concentration ([Hb]) and thus measures the same hemodynamic response as functional magnetic resonance imaging (fMRI), but is portable, non-confining, relatively inexpensive, and is appropriate for long-duration monitoring and use at the bedside. Like fMRI, it is noninvasive and safe for repeated measurements. Patterns of [Hb] changes are used to classify cognitive state. Thus, fNIRS technology offers much potential for application in operational contexts. For instance, the use of fNIRS to detect the mental state of commercial aircraft operators in near real time could allow intelligent flight decks of the future to optimally support human performance in the interest of safety by responding to hazardous mental states of the operator. However, many opportunities remain for improving robustness and reliability. It is desirable to reduce the impact of motion and poor optical coupling of probes to the skin. Such artifacts degrade signal quality and thus cognitive state classification accuracy. Field application calls for further development of algorithms and filters for the automation of bad channel detection and dynamic artifact removal.
This work introduces a novel adaptive filter method for automated real-time fNIRS signal quality detection and improvement. The output signal (after filtering) will have had contributions from motion and poor coupling reduced or removed, thus leaving a signal more indicative of changes due to hemodynamic brain activations of interest. Cognitive state classifications based on these signals reflect brain activity more reliably. The filter has been tested successfully with both synthetic and real human subject data, and requires no auxiliary measurement.
This method could be implemented as a real-time filtering option or bad channel rejection feature of software used with frequency domain fNIRS instruments for signal acquisition and processing. Use of this method could improve the reliability of any operational or real-world application of fNIRS in which motion is an inherent part of the functional task of interest. Other optical diagnostic techniques (e.g., for NIR medical diagnosis) also may benefit from the reduction of probe motion artifact during any use in which motion avoidance would be impractical or limit usability.
This work was done by Angela Harrivel and Tristan Hearn of Glenn Research Center.
Inquiries concerning rights for the commercial use of this invention should be addressed to
NASA Glenn Research Center
Innovative Partnerships Office
Attn: Steven Fedor
Mail Stop 4–8
21000 Brookpark Road
Cleveland
Ohio 44135.
LEW-18952-1
This Brief includes a Technical Support Package (TSP).
Functional Near-Infrared Spectroscopy Signals Measure Neuronal Activity in the Cortex
(reference LEW-18952-1) is currently available for download from the TSP library.
Don't have an account?
Overview
The document is a Technical Support Package from NASA's Glenn Research Center, focusing on Functional Near-Infrared Spectroscopy (fNIRS), a promising neurological sensing technique. fNIRS is particularly relevant for optimizing human performance in transportation operations, such as commercial aviation, by measuring neuronal activity in the cortex.
The primary goal of fNIRS is to assess cognitive states through the pattern classification of functional activations. This technique quantifies changes in hemoglobin concentration in the brain based on optical intensity measurements. The document emphasizes the importance of developing algorithms and filters to remove dynamic artifacts that can interfere with the accuracy of fNIRS signals.
A significant contribution of the document is the introduction of a novel adaptive filter method designed to improve the quality of fNIRS signals in real-time. This method utilizes the frequency domain phase shift signal to tune a Kalman filter, which has been shown to effectively reduce unwanted motion artifacts by at least 43% and enhance the contrast of the filtered oxygenated hemoglobin signal by over 100%. This advancement is crucial for applications where motion is inherent, such as in aviation, where maintaining signal integrity is vital for accurate cognitive state monitoring.
The document also discusses previous work that has improved the comfort of optical head probes, allowing for extended monitoring periods of over an hour. It reports an average accuracy of 70% for real-time classification of attentional states using Support Vector Machines, based on artifact-free training data. This indicates a significant step forward in the practical application of fNIRS for attentional monitoring.
Overall, the document highlights the potential of fNIRS as a non-invasive tool for real-time monitoring of cognitive states, which could significantly enhance safety and performance in high-stakes environments like aviation. It underscores the ongoing efforts to refine the technology, improve data processing methods, and adapt the system for various cognitive state detections, making it a valuable resource for researchers and practitioners in the field of neuroscience and human factors engineering.
