Real-time thermoacoustic imaging and thermometry in bovine udder tissue comparing two calibration methods

Thermoacoustic (TA) thermometry requires calibration to establish the relationship between the change in TA signal with the change in temperature. Previous studies have used a priori slopes values (i.e. TA signal vs temperature), typically determined experimentally from previously performed standard calibrations. However, standard calibration requires more than one test and is not robust to changing sample orientation or composition. The goal of this study is to compare a previously studied self-calibration method to the standard calibration method in estimating temperature using real-time 3D TA thermometry in bovine udder tissue kept in the same orientation for both methods. The self-calibration and standard methods estimated temperatures of the udder sample during localized heating with a normalized root mean square errors (NRMSE) of 7.1% and 8.4%, respectively. The predicted temperatures produced by the self-calibration method were less variable than the standard method. Overall, the qualitative and quantitative results comparing predicted change in temperature suggested that self-calibration is as accurate as standard calibration with less variability and less experimentation. These advantages point to self-calibration being a more preferred method to provide temperature predictions for feedback control in future breast cancer ablation therapy.