Type2 fuzzy set based automatic shadow region segmentation in intra vascular ultrasound images

This paper presents an automated algorithm for shadow regions detection in IVUS images using a histogram modification method for threshold selection. The method employs the Type2 Fuzzy Set together with spatial information for shadow region separation from the background image. Shadow regions usually occur behind the calcium deposits making it difficult or impossible to process tissue images in these regions. Results show that the proposed method is effective in detection of parts of the image in the shadow regions. Applying to a data set of 40 IVUS frames the results show a good agreement between the automatically detected and manually defined shadow regions boundaries by the proposed method and the expert, respectively.