对小尺度生物体热声成像的反卷积重建法

Deconvolution Reconstruction of Thermoacoustic Imaging for Small-scale Objects

目的为了提高热声成像的速度，针对小尺度待测生物体提出一种基于反卷积的热声成像算法：反卷积重建法。方法首先通过探测到的声压函数构造出一个新函数，然后基于反卷积方法从此新函数重建出待测生物体内部的电磁波吸收系数的分布。通过计算机仿真实验比较反卷积重建法和当前流行的时域重建法与滤波反投影法的性能。结果对于小尺度的待测生物体，反卷积重建法的精度与时域重建法相当，略好于滤波反投影法。在仿真实验条件下，反卷积重建法的速度约是时域重建法的4—6倍、滤波反投影法的25—100倍。结论对于小尺度生物组织的热声成像问题，反卷积重建法是一种快速有效的算法。

Objective To increase the speed of thermoacoustic imaging, a deconvolution reconstruction （DR） algorithm is proposed for small-scale need detecting living beings, on the basis of deconvolution of thermoacustic imaging algorithm. Methods A new function was firstly constructed from the detected acoustic pressure function. Then, the distribution of electromagnetic wave absorption coefficient of the need detecting living beings could be reconstructed from this new function based on the deconvolution method. Computer simulation studies were carried out to compare the DR algorithm with the two popular algorithms： the time-domain reconstruction （TDR） and the filtered back projection （FBP）. Results For small-scale need detecting living beings, the accuracy of DR was nearly equivalent to TDR and slightly better than FBP. However, DR was about 4 to 6 times faster than TDR and about 25 to 100 times faster than FBP under the simulation experiment condition. Conclusion DR is a fast and effective algorithm of thermoacoustic imaging for small-scale living beings.