Hybrid Neural Network-Based Fuzzy Inference System Combined with Machine Learning to Detect and Segment Kidney Tumor
P Srinivasa Rao, Pradeep Kumar Bheemavarapu, D Swapna, Subba Rao Polamurı, M. Madhusudhana Subramanyam · 2023
In the current technological era, it is essential to identify diseases at the preliminary stages so as to lead a healthy life. One of the typical diseases in adults is renal tumor, also called kidney cancer. Through early diagnosis of tumor, the possibilities of treatment and survival rate of the patients can be improved. Existing techniques include blood tests and CT scan analysis by nephrologists to detect tumors, which is tedious. Hence, an automated kidney tumor detection system needs to be developed for early detection. Many methods were proposed depending on deep learning (DL) and machine learning (ML) techniques by the researchers. With the advancements in research methodologies, an attempt is made to develop a hybrid neural fuzzy inference system combining convolutional neural networks (CNN) with machine learning (ML) techniques, which is explored for detection and segmentation of tumors in kidneys. The proposed hybrid system to detect and segment kidney tumors is compared to traditional and competitive algorithms and is proven to be accurate in segmentation with 98.69% classification accuracy and 0.8923 dice coefficient for tumors.