Evolutionary Deep Learning Based Liver Tumour Segmentation Optimized With Pso In Ct Scans

Abstract

Image processing and computer vision are the primary research areas in today ‘s world.  This visible spectrum is limited to human eye due to the size, distance, structure, lighting characteristics of the object. Hence, image processing help us to visualize all the objects including all the characteristics. Numerous applications are based on the concept of image processing based on certain parameters such as improvement in imaging, use of requirement of applications and etc.,in the research area of medical diagnosis and treatment planning image processing plays vital role. Liver tumour segmentation from the CT scans is still a challenging problem. Evolutionary computing with deep learning are the current research that will used in many applications high satisfaction. In this work, deep learning based segmentation using convolutional deep neural network optimized with PSO has been used for segmentation and extraction of liver tumour. This work consists of four steps such as image pre-processing, liver localization using standard CDNN model, liver segmentation using the proposed CDNN with PSO and tumour segmentation