Algorithms for reconstructing three-dimensional temperature distributions inside an object from its intrinsic thermal acoustic radiation are studied. A regularization method for solving the inverse problem, taking into account the specific physical behavior of the temperature distribution inside the human body, is proposed. A comparison between results of reconstruction obtained by different methods reveals that, using the Fourier transform in combination with algorithms of signal and pattern reconstruction modeled with the help of differ ential equations, one can construct a passive thermograph with a rather low rms error and high spatial resolu- tion. In this technique, the major pan of the error is determined by the reconstruction algorithm rather than by the accuracy of the physical instrument.
展开▼
