A Detection and Segmentation of Medical Image Using Machine Learning Algorithms

Abstract

Steatosis, often known as fatty liver disease, is a common disorder brought on by an overgrowth of fat in the liver. There is a modest amount of fat in a healthy liver. A problem arises when fat makes up 5% to 10% or more of the weight of our liver. This illness is linked to a high risk of morbidity and mortality. The goal of our research proposal is to create a machine learning model that can help physicians identify high-risk patients, provide a distinct diagnosis, and prevent and treat FLD. 55 of the 105 people who were recruited had their FLD assessed. The sickness was found using classifier models. Among the models available are decision trees, K-Neighbors classifiers, logistic regressions (LR), random forests (RF), and linear regressions and Decision Tree Classifier. The link between the model's actual and anticipated values was examined using the confusion matrix. According to the study's findings, machine learning algorithms can accurately predict FLD. According to our research, the Logistic Regression model is the most reliable regression model. The Random-Forest-Classifier has a higher accuracy of 68 percent than the other classification models. Our method could be used to find FLD patients who could have a big impact on how treatments are delivered. The use of this algorithm for early prediction could lower medical expenses and save money on treatment.