Comparison of machine learning algorithms in the context of deep diagnosis of endometrial carcinoma with myometrial invasion

Endometrial cancer is one of the most common primary malignant tumors in women. The depth of myometrial invasion is an independent risk factor for the prognosis of endometrial cancer, and it is also an important basis for clinical staging by the International Federation of Obstetrics and Gynecology. Therefore, it is very important to accurately assess the depth of myometrial invasion of endometrial cancer before surgery. As a routine inspection method for the depth of myometrial invasion of endometrial cancer, ultrasonography has the characteristics of short time, low price but low accuracy. Combined with the routine examination indicators and ultrasound data of patients with endometrial cancer, this paper uses machine learning algorithms such as logistic regression, decision tree, random forest, clustering, multilayer perceptron and other algorithms to classify and predict data samples. The results show that random forest showed the best predictive analysis performance with prediction accuracy, sensitivity and specificity of 97.7%, 94.1% and 100.0%, respectively.