A compressive sensing approach for progressive transmission of images

An interactive and adaptive joint source-channel coding for progressive transmission of medical images is proposed. A hierarchical alternative least squares (HALS) based compressive sensing (CS) is effectively used for adaptive selection of the wavelet coefficients and nonlinearly thresholding the transform coefficients. In this modality the source compression rate is influenced by the proximity to the region of interest (RoI), which often includes significant diagnostic information. Also, the channel coding scalability is affected by both the proximity to the RoI and the channel characteristics. The results are later compared with the well-established embedded zero-tree wavelet (EZW) approach in terms of compression rate and efficiency.