An efficient algorithm and architecture for medical image segmentation and tumour detection

Abstract

Medical image segmentation is very important for radiotherapy planning and cancer diagnosis. There are many techniques for medical image segmentation based on thresholding, classification, and multiresolution analysis (MRA). This paper proposes a system based on MRA and artificial intelligence techniques (AI) for tumour segmentation in DICOM images. The slowest parts of the proposed system have been accelerated using field programmable gate arrays (FPGA). Hardware implementation of Haar wavelet transform based factorization approach (HWTF) on reconfigurable hardware using distributed arithmetic (DA) principles is presented. The developed architecture can be integrated into a system for automatic detection and segmentation of tumour in positron emission tomography (PET) images.