A Dual Channel Multiscale Convolution U-Net Methodfor Liver Tumor Segmentation from Abdomen CT Images
Joel M. Dickson, Arul Lincely, Alice Nineta · 2022 International Conference on Sustainable Computing and Data Communication Systems (ICSCDS) · 2022
Nowadays, the automatic segmentation of liver tumor is needed for various medical applications like pathological detection of liver diseases, surgical planning, and postoperative evaluation. The liver is a massive, meaty organ on the right part of the abdomen that plays an important part in our digestive system. Cancerous tumors in the liver can cause a serious threat to human life. Currently, automatic liver tumor segmentation has a lot of problems, such as complex tissues, fusion of features of the same scale, insufficient learning of features, and irregular shapes of tumors. Typically, more networks are used to learn the important features, and fuse the features of same scale between encoders and decoders. But still, segmentation of tumor is one of the main challenges in the liver. Besides, soft organ removal, noise removal, and edge detection are important stages in the segmentation of liver tumor. Aiming at these problems, this paper introduced a Dual Channel Multiscale Convolution Unet (DCMC-Unet) to segment the liver tumor. An Intensity based threshold method (IBTM) is used for removing the soft organs, and Z-score normalization is utilized to eliminate the noise. The Enhanced sobel method (ESM) is used to enhance the local features of the liver. Finally, the enhanced features are fed to the DCMC-Unetmodel to segment the tumor from the liver. Experimental analysis illustrates that the DCMC-Unet approach attains higher segmentation results than existing models in terms of several qualitative metrics.