Detection and Classification of Liver Cancer using Neural Networks
Diagnosisof Cancer and treating have a great importance due to many prevalent stages of diseases, most of the people now a days died because of cancer effected organs like lung, liver, brain, oral etc. Liver cancer is generally analysed by three diverse tests like blood test, image test and biopsy. To facilitate the task in a very short time in an effective and efficient process forthat, we used various computer-aided diagnostic systems to identify liver cancer. MRI, CT and Ultrasound scanned images are used for detecting tumor. The proposed methodology uses CT scanned images. Initially we have taken a scanned image as input and then perform pre-processing then that image is ready for processing. Next step is to improve and enhance the quality of the input image for that we perform image enhancement and the technique used for that is Histogram equalization and after that perform image segmentation with the use of Thresholding. The image is filtered using Sobel’s filter i.e. applying mask and then perform morphological image processing i.edilation takes place after that feature extraction is carried out. Finally, the classification of liver cancer is done by using neuro classifier (CNN) by using extracted parameters like entropy, contrast and energy functions of the image with better accuracy and performance. The proposed methodology reduces the time complexity and computational complexity.