Functional Near Infrared Spectroscopy: from a single-channel towards a multichannel approach

N.J.M. Colier, M.C. van der Sluijs, J. Menssen and B. Oeseburg

Department of Physiology, University Medical Center St. Radboud, Nijmegen, The Netherlands

Introduction
During the last six years, single and dual channel Near Infrared Spectroscopy (NIRS) has demonstrated that changes in brain activitity can be assessed non-invasively. Similar to functional magnetic resonance imaging (fMRI), NIRS monitors changes in concentration of deoxyhemoglobin ([HHb]). However, NIRS is also capable to provide continuous and direct information on oxyhemoglobin ([O2Hb]) and on total hemoglobin ([tHb]). Compared to fMRI, fNIRS has a faster sampling rate, is much less sensitive to movement artifacts, portable and relatively cheap. An inherent disadvantage of NIRS is the lower spatial resolution compared to fMRI. This can largely be overcome by using a multi-channel device that can image (part) the human brain. Our goal was to develop a highly portable and sensitive optical tomography device with fast sampling rate. Parallel to this process we developed software, using Matlab, to provide the user with the tools for direct and fast data analysis.

Materials and methods
The instrument consists of 4 light sources (laser diode) and 3 detectors (avalanche photodiode). The technology of the device is based on the continuous wave Oxymon® device, developed by the Physiology and Biomedical Enigineering Department of the UMC St. Radboud. Every light source has two wavelengths: 775 and 850 nm. Detectors and light sources are coupled to optical fibers. The fibers of 2 detectors are split, to obtain a total of 5 detectors. Detectors and light sources are placed alternately on square lattice points. The distance between the point is fixed at either 4.0 or 3.5 cm. Detectors are placed on the corners of the square and one in the middle of the square. The total imaging area is 8 by 8 or 7 by 7 cm. Currently two desktop PCs are used to run the instrument. Additionally to the hemoglobin signals, up to 16 anolog inputs can be stored simultaneously. The fastest sampling rate of the imager is 50 Hz. Using this high sampling rate the imager does not only measure changes in concentration of [HHb], [O2Hb] and [tHb], but also measures cerebral arterial saturation.

Results
The functionality of the instrument was tested using a liquid phantom (20% milk in water) with a rotating rod of 1 cm in diameter and on the human brain during a finger-tapping task (20 s stimulus duration, tapping rate 2 Hz, right hand). A picture with time frames of [O2Hb] and [HHb] during the finger-tapping task is shown in Figure 1.

 

Figure 1. Top: [O2Hb]. Bottom: [HHb].

Conclusion
The optical imager provides promising results for its application in human cortex mapping.

Acknowledgements
We would like to thank Prof. Ferrari and Dr. Quaresima for their stimulating discussions on optical topography and their help with the testing of the instrument.

This research has been supported in part by the European Commission-DGXII and the Dutch Technology Foundation.


Poster presented at Measuring Behavior 2000, 3rd International Conference on Methods and Techniques in Behavioral Research, 15-18 August 2000, Nijmegen, The Netherlands

© 2000 Noldus Information Technology b.v.