Permanent URL to this publication: http://dx.doi.org/10.5167/uzh-17583
Spichtig, S; Hornung, R; Brown, D W; Haensse, D; Wolf, M (2009). Multifrequency frequency-domain spectrometer for tissue analysis. Review of Scientific Instruments, 80 (2):024301 .
View at publisher
In this paper we describe the modification and assessment of a standard multidistance frequency-domain near infrared spectroscopy (NIRS) instrument to perform multifrequency frequency-domain NIRS measurements. The first aim of these modifications was to develop an instrument that enables measurement of small volumes of tissue such as the cervix, which is too small to be measured using a multidistance approach. The second aim was to enhance the spectral resolution to be able to determine the absolute concentrations of oxy-, deoxy- and total hemoglobin, water, and lipids. The third aim was to determine the accuracy and error of measurement of this novel instrument in both in vitro and in vivo environments. The modifications include two frequency synthesizers with variable, freely adjustable frequency, broadband high-frequency amplifiers, the development of a novel avalanche photodiode (APD) detector and demodulation circuit, additional laser diodes with additional wavelengths, and a respective graphic user interface to analyze the measurements. To test the instrument and algorithm, phantoms with optical properties similar to those of biological tissue were measured and analyzed. The results show that the absorption coefficient can be determined with an error of <10%. The error of the scattering coefficient was <31%. Since the accuracy of the chromophore concentrations depends on the absorption coefficient and not on the scattering coefficient, the <10% error is the clinically relevant parameter. In addition, the new APD had similar accuracy as the standard photomultiplier tubes. To determine the accuracy of chromophore concentration measurements we employed liquid Intralipid® phantoms that contained 99% water, 1% lipid, and an increasing concentration of hemoglobin in steps of 0.010 mM. Water concentration was measured with an accuracy of 6.5% and hemoglobin concentration with an error of 0.0024 mM independent of the concentration. The measured lipid concentration was negative, which shows that the current setup is not suitable for measuring lipids. Measurements on the forearm confirmed reasonable values for water and hemoglobin concentrations, but again not for lipids. As an example of a future application, chromophore concentrations in the cervix were measured and comparable values to the forearm were found. In conclusion the modified instrument enables measurement of water concentration in addition to oxy- and deoxyhemoglobin concentrations with a single source-detector distance in small tissue samples. Future work will focus on resolving the lipid component.
1 download since deposited on 30 Mar 2009
1 download since 12 months
|Item Type:||Journal Article, refereed, original work|
|Communities & Collections:||04 Faculty of Medicine > University Hospital Zurich > Clinic for Neonatology|
|Dewey Decimal Classification:||610 Medicine & health|
|Deposited On:||30 Mar 2009 09:17|
|Last Modified:||20 May 2016 10:13|
|Publisher:||American Institute of Physics|
|Free access at:||Publisher DOI. An embargo period may apply.|
Users (please log in): suggest update or correction for this item
Repository Staff Only: item control page