Machine Learning Based Earlier Identification of Liver Disease Using Ultrasound Images

Liver disease begins in the ovaries and is especially hazardous for women. As a result, abnormal cells can potentially spread to other organs. Liver disease is a type of dangerous development that affects the ovaries in females, and it is difficult to detect at an early stage, which is why it remains one of the leading causes of death from disease. Unquestionable confirmation of intrinsic and natural components is crucial for the development of novel systems to detect and eliminate hazard. We propose the detection and categorization of liver illness in ultrasound pictures using hybrid machine learning techniques. Using Machine Learning Algorithms and a vote-based classifier to classify liver tissues as fatty or normal based on ultrasound picture attributes and voting. Our created method provides four major contributions: first, the classification of liver pictures as normal or fatty is performed without a segmentation phase. Secondly, compared to our suggested effort, the datasets in earlier works were inadequate. The third contribution is a collection of 26 characteristics. On the basis of the presented methods, sigmoid radial basis function neural network (SRBFNN) and Gray-Level Co-Occurrence Matrix are the recovered features (GLCM). The fourth contribution is the voting classifier that is utilised to determine the type of liver tissue. Fruit Fly Optimization concludes the categorization process (FFO). We have achieved a classification accuracy of 98.24% using our proposed approach. Convolutional Neural Network (CNN), Support Vector Machine (SVM), K-Nearest Neighbour (KNN), Random Forest (RF), and Naive Bayes are compared with our proposed system (NB). The experimental results demonstrate that the classification accuracy of our suggested system is superior to that of the leading classification strategies.