In this work, we aim to obtain our designedsystem performance in order to transmit MPEG-4 medicalimages over indoor wireless channels. These kinds of imageslook similar regardless of their subject. Thus, the sourcetransition probability matrix (STPM) used in a MAP decodingprocess can be computed and configured at the Viterbireceiver in advance. Moreover, to overcome the burst noisetransmission problem, we employ a ring convolutional encoderwith reconfigurable random inter-leavers. The systemperformances are summarized, based on the transmission overa Nakagami-m channel with block-fading parameter Ns. Thus,the modulating signals are not only corrupted by additivewhite Gaussian noise but also distorted by the multiplicativeand correlative fading variables. The STPM and residualredundancy of both “Ultrasound” and “CTAnkle” images aredescribed and computed. The simulation results of both“Ultra-sound” and “CTAnkle” images are summarized for m=1, 2 and 4with Ns = 10 and 20