Towards personalized and versatile monitoring of temperature fields within heterogeneous tissues during laser therapies

Abstract

Advancements in medical laser technology have paved the way for its widespread acceptance in a variety of treatments and procedures. Selectively targeting particular tissue structures with minimally invasive procedures limits the damage to surrounding tissue and allows for reduced post-procedural downtime. In many treatments that are hyperthermia-based, the efficiency depends on the achieved temperature within the targeted tissues. Current approaches for monitoring subdermal temperature distributions are either invasive, complex, or offer inadequate spatial resolution. Numerical studies are often therapy-tailored and source tissue parameters from the literature, lacking versatility and a tissue-specific approach. Here, we show a protocol that estimates the temperature distribution within the tissue based on a thermographic recording of its surface temperature evolution. It couples a time-dependent matching algorithm and thermal-diffusion-based model, while recognizing tissue-specific characteristics yielded by a fast calibration process. The protocol was employed during hyperthermic laser treatment performed ex-vivo on a heterogeneous porcine tissue, and in-vivo on a human subject. In both cases the calibrated thermal parameters correlate with the range of values reported by other studies. The matching algorithm sufficiently reproduced the temperature dynamics of heterogeneous tissue. The estimated temperature distributions within ex-vivo tissue were validated by simultaneous reference measurements, and the ones estimated in-vivo reveal a distribution trend that correlates well with similar studies. The presented method is versatile, supported by the protocol for tissue-specific tailoring, and can readily be implemented for temperature monitoring of various hyperthermia-based procedures by means of recording the surface temperature evolution with a miniature thermal camera implemented within a handheld laser scanner or similar.

Fig. 1.

The tissue-tailored approach is enabled by the TPE protocol, where tissue thermal data is acquired by fitting a simulated tissue response to the one measured with a thermal camera. The estimation algorithm links the TPE outcome and yields a personalized, time-dependent database by calculating the evolutions of arbitrarily generated STDs. Finally, the STD within the tissue is estimated by correlating the T_s evolution during the relaxation period with the ones clustered in a database.

Fig. 2.

Measured T_s evolution (colored lines) during the heating and relaxation cycle of the TPE process: The closest fitting emulations from the digital model are shown with black dashed lines. Three different average intensities were used in the heating phase I₁=0.08, I₂=0.1 and I₃=0.12 W/cm². Graph a. shows the T_s dynamic on an ex-vivo tissue sample, and b. the T_s dynamic for an in-vivo case. Graphs b. and c. show influence of blood perfusion rates on goodness of fit (RSS) between TPE model and measured T_s dynamic for ex-vivo and in-vivo tissues respectively.

Fig. 3.

Fitting of T_s evolution and prediction of the STD in porcine tissue: Red color denotes CC1, orange CC2 and blue CC3. Graphs a. and c. show the measured surface temperature (colored lines) and fits (black lines) for Nd:YAG and Alexandrite, respectively. Graphs b. and d. show comparison between the measured STD (colored lines) and estimated distribution (black lines) for Nd:YAG and Alexandrite, respectively.

Fig. 4.

Fitting of T_s evolution and prediction of the STD within a human subject: Red color denotes CC1, orange CC2 and blue CC3. Graphs a. and c. measured surface temperature (colored lines) and fits (black lines) for Nd:YAG and Alexandrite, respectively. Graph b. shows the estimated STDs for Nd:YAG, together with the MCML outcome for the irradiation time of ∼80 s. Graph d. Shows STD estimations for an Alexandrite-based HTLT.