Applicability of genetic algorithms to reconstruction of projected data from ultrasonic tomography

The use of a-priori information, where available, is an important step in solving an already computationally expensive tomographic imaging problem [1]. Here, an enhanced genetic algorithm based reconstruction technique is proposed that is capable of detecting the shape, size and location of multiple types of inclusions of known physical properties in a given test specimen. Preliminary results are found to be better than those reported with MART1. Simulations show that the algorithm is consistent for a wide range of grid sizes and geometries of inclusion(s). A logarithmic time complexity analysis gives a linear relationship between number of unknowns and reconstruction times, thus establishing the predictability of the algorithm.