Abstract
The potentiality of passive thermoacoustic tomography (PTT) is considered theoretically. The problem is formulated to reconstruct the distribution of the thermodynamic temperature at a depth from the acoustic radiation generated by thermal noise and measured by a set of piezotransducers on the body surface. The 2-D inverse problem has been studied. Three mathematical methods of reconstruction were investigated: (1) the least squares method; (2) Tikhonov's regularization method; and (3) the method of elimination of non-physical solutions. The reconstruction in square as well as in rectangular areas was studied. If one uses the least squares method the square area can be separated into 5 5 subareas, the typical dimension of every one is 0.5-2 cm for measurements conducted in the frequency range 2-0.5 MHz. The precision deltaT in temperature reconstruction in different layers is about 0.1-0.2 K for smooth distributions of the thermodynamic temperature and 0.1-0.4 K for abrupt distributions, if the number of different scans is about 102 and the time of data collecting is about 1 min.