Research summary: Application of a temperature-dependent fluorescent dye (Rhodamine B) to the measurement of radiofrequency radiation-induced temperature changes in biological samples

Yuen Yuen Chen and Andrew Wood

Faculty of Life and Social Sciences, Cellular Neuroscience Laboratory, Brain Sciences Institute, Swinburne University of Technology, Hawthorn, VIC, Australia


Summary of research published in Biolelectromagnetics, Vol. 30, No. 7, pp 583-590.

The computed Specific Absorption Rate (SAR) in both in-vitro and in-vivo preparations has often been used as an indicator of whether observed radiofrequency (RF) effects can be due to a thermal or a non-thermal mechanism. However, because of the high variation in SAR in many of these preparations, the diffusion or convection of heat makes the local changes in temperature difficult to estimate. This is particularly true in complex tissue samples such as brain slices or in whole brain.  Although it is possible to penetrate tissue with fluoroptic probes, this produces damage and can only measure from a limited number of locations.

The method described in this paper involves a commonly-used fluorescent dye, Rhodamine B, which is readily taken up by tissue and whose fluorescent intensity is markedly decreased as temperature rises. The intensity is mapped in 3D by confocal microscopy, which, in essence, produces a ‘stack’ of 2D ‘slices’ at successive depths within the sample. By taking images at various times before, during, and after RF exposure, the change of temperature at various locations in the sample was followed.  The paper also reports on initial calibration of the dye in non-living or fixed samples against temperature estimated by fluoroptic probes. From this calibration, the fluorescent intensity fell by 3.4+/-0.2 % per degree Celsius. The RF energy was supplied to a purpose-built exposure chamber in which simultaneous confocal images are obtained and in which the SAR distribution has previously been extensively modeled. Changes in fluorescence reveal a temperature gradient within the sample (which was approximately 0.6 x 0.6 x 0.1 mm, with a ‘stack’ of 15 x 2D images each 6.5 microns thick), which was not anticipated from SAR measurements.

We anticipate that this work will contribute to a better understanding of SAR in relation to localized changes in temperature in biological samples.  Ongoing work is examining the use of Rhodamine B in fresh (non-fixed) tissue to determine the initial rate of temperature rise in the first few seconds after the RF power has been turned on compared with what is predicted by thermal modeling.