Murmur Separation and Classification from Heart Sound

The goal of this study was to use patient information and data from preoperative exams to predict the pathological diagnosis of ovarian cancers using artificial intelligence (AI). This investigation was conducted using a variety of current algorithms, including both classic and more recent ones that were found in the specialized literature. Our primary goal was to identify medical problems quickly and map them onto compact tabular data sets to improve the effectiveness of early disease identification. The data used in this current study includes 349 Chinese patients with a total of 49 variables, including demographics, common blood tests, general chemistry, and tumor markers [1]. The test set consists of 114 patients (89 with benign ovarian tumors and 25 with ovarian cancer). In this review study, we used K-Nearest Neighbor, Naïve Bayes, Random Forest, Decision Trees, Support Vector Machines, Logistic Regression, and TabPFN algorithms.