Optimization of a random linear ultrasonic therapeutic array based on a genetic algorithm

Given their advantage of suppressing grating lobes, randomly arranged linear arrays have potential for use in ultrasonic treatment. The current work proposes a method based on genetic algorithm to optimize the random arrangement of array elements, so that the suppression effect of grating lobes can be significantly improved with reduced calculating time. The maximum and average kerfs of array elements are used as genes, and the ratio of the maximum to the secondary maximum sound pressure at the focal plane is used as the optimized target. Typically, the calculation requirements of the current method can be reduced to ∼ 25% of the traversing method. We further discuss how the kerf width, the effective ratio of element areas and the ratio of focal distance to array aperture affect the suppression of grating lobes. For a typical linear array with 32 elements (1-MHz operating frequency, 1.5-mm element width and 150-mm focal distance), the results suggest that the grating lobes are suppressed well when (1) the ratio of maximum width to average width of the element is between 5 and 8, (2) the ratio of the effective element area to the area of the whole array is between 0.5 and 0.9, and (3) the ratio of the effective emission aperture to the actual emission aperture of the array is as large as possible. Based on optimized parameters, an experimental array was fabricated and the measured results of corresponding sound field were entirely consistent with the simulated results (Given her role as an Associate Editor of this journal, Juan Tu had no involvement in the peer-review of articles for which she was an author and had no access to information regarding the peer-review. Full responsibility for the peer-review process for this article was delegated to another Editor of this journal.).