A flexible algorithm for layout-optimized sparse cMUT arrays

Optimization of array layouts for ultrasonic imaging has been a research topic for several decades. The task is to find layouts giving optimal image quality with a reduced number of active elements. The research is motivated from the ability to reduce the cost and complexity of ultrasonic systems. Recently, cMUT array technology has made possible the production of 2D arrays with large flexibility on element size, shape and position. For such arrays an optimal number, form and placement of elements would give a minimal cost system with an optimal image quality. In this paper, a fast and flexible simulated annealing algorithm for array layout optimization is presented. In calculation of the array response, the shape, orientation and position in 3D space of each element, and the focus point of the array are accounted for. For each iteration, the array response is calculated using a differential update based on the response in the previous iteration. Potential improvements from using layouts other than the standard linear grid with square elements are investigated. Two designs of possible cMUT arrays with minimal sidelobe level are presented. The first is a sparse ring array having increasing element sizes for increasing ring radii. The second is a hex-grid optimized array with hexagonal and circular elements. The designs are compared to a corresponding linear grid design with square elements.