Sustainable method of automatic detection of tumor using super pixel segmentation

Abstract

Liver cancer is the leading cause of cancer-related death worldwide. Since the radiologist's ability to diagnose liver cancer at an early stage is zero, the prognosis is poor. According to numerous investigations performed so far, the nodule segmentation algorithms are clearly ineffective. As a result, for specific pulmonary nodule segmentation, this study made use of the advanced optimization tool and centralized super pixels segmentation based iterative clustering (SSBIC). To remove noise from the images, start by using ADF and unsharp masking enhancement techniques. In order to predict abnormal liver tissue, an enhanced nodule image sequence is subjected to the Super pixel Segmentation Based Iterative Clustering (SSBIC) algorithm. Finally, to photograph liver nodules, a deep learning-based Advanced GWO with CNN (AGWO-ONN) and an Advanced GWO with ONN (AGWO-ONN) are used (AGWO-CNN).For nodule slice order, the average segmentation time is 1.06s. The classification accuracy of the Advanced GWO with ONN (AGWO-ONN) method is 97 percent, while the classification accuracy of the Advanced GWO with CNN (AGWO-CNN) method is 97.6 percent.