P1B-5 Real-time Tissue Compression Quality Feedback for Optimized Freehand Elasticity Imaging

As pathological conditions often produce changes in biological tissue stiffness, real-time freehand elasticity imaging provides valuable information, facilitating the detection and diagnosis process. Its disadvantage, however, consists of exhaustive operator training, as obtaining superior quality strain images requires continuous adjustment of the compression technique. The sonographer needs to maintain a constant compression rate, exclusively on the axial direction of the imaging transducer, while avoiding lateral and out-of-plane tissue motions. A compression analysis algorithm was designed to estimate in real-time the strain image quality prior to actually computing the image by assessing the quantity and quality of tissue compression versus depth. The algorithm employed a criterion for automatic selection of the most advantageous frame pairs (pre- and post-compression) producing elasticity images of optimal dynamic ranges and signal-to-noise ratios while providing live graphical compression feedback to the operator. The algorithm was implemented on a modified commercial scanner (Aloka SSD5500) equipped with a newly developed static elasticity imaging module. The strain images were displayed using a graphic interface specially developed to include both the compression feedback and a combined B-Mode / strain display. Results of breast elasticity were obtained at Yale New Haven Hospital, CT for various solid masses and cysts, including benign and malignant cases which validated the algorithm. Over 30 patients were scanned during this initial phase. The operator training was facilitated by the provided graphical compression feedback significantly reducing the learning period. The frame pair selection criterion optimized the trade-off between the amount of images obtained and their quality and reduced radically the amount of artifact in the images while lowering the computational burden. During the offline review process of the collected elasticity data, the information indicating the tissue compression quantity and quality confirmed the quality of raw data behind the elasticity images