Monitoring of deep brain temperature in infants using multi-frequency microwave radiometry and thermal modelling

In this study we present a design for a multi-frequency microwave radiomete r aimed at prolonged monitoring of deep brain temperature in newborn infant s and suitable for use during hypothermic neural rescue therapy. We identif y appropriate hardware to measure brightness temperature and evaluate the a ccuracy of the measurements. We describe a method to estimate the tissue te mperature distribution from measured brightness temperatures which uses the results of numerical simulations of the tissue temperature as well as the propagation of the microwaves in a realistic detailed three-dimensional inf ant head model. The temperature retrieval method is then used to evaluate h ow the statistical fluctuations in the measured brightness temperatures lim it the confidence interval for the estimated temperature: for an 18 degrees C temperature differential between cooled surface and deep brain we found a standard error in the estimated central brain temperature of 0.75 degreesC . Evaluation of the systematic errors arising from inaccuracies in model pa rameters showed that realistic deviations in tissue parameters have little impact compared to uncertainty in the thickness of the bolus between the re ceiving antenna and the infant's head or in the skull thickness. This highl ights the need to pay particular attention to these latter parameters in fu ture practical implementation of the technique.