Dynamic deep temperature recovery by acoustic thermography using neural networks

Abstract

In an experiment, the deep temperature, which changed with time, was recovered for a model object, bovine liver. The liver was heated for 6 min by laser radiation (810 nm), transmitted via a light guide to a depth of 1 cm. During heating and subsequent cooling, the deep temperature was measured by acoustic thermography. For independent control, we used three electronic telemeters, the indications of which were also subsequently recovered. Deep temperature was recovered using a neural network with a time delay. During the last 2 min of heating, the mean square error of recovery for an averaging time of 50 s did not exceed 0.5°C. Such a result makes it possible to use this method for solving a number of medical problems.