Predicting the stage of gastric cancer after gastrectomy based on machine learning algorithms

Abstract Background Gastric cancer (GC) is the fourth most common cause of cancer death worldwide, with a 5-year survival rate of less than 40%. One of the most important methods for diagnosing stomach cancer is endoscopy, which is quite costly and invasive. The aim of this study was to develop machine learning-based diagnostic prediction models for the stage of GC. Objectives To create a highly accurate predictive model for the stage of GC in patients via a noninvasive method based on machine learning (ML). Methods In this study, data from 996 patients with GC after gastrectomy were utilized. The data were split into groups, trained and tested, and a ratio of 8:2 was used to develop different machine learning models. Furthermore, the six different machine learning algorithms used in predicting the stage of GC include decision tree (DT), K nearest neighbor (KNN), logistic regression (LR), naive Bayes (NB), random forest (RF), and support vector machine (SVM) methods. Results: The analysis of the demographic variables revealed statistically significant differences in the PLR and NLR and other parameters between the two groups of patients with stages I and III gastric cancer (P < 0.05). Results The analysis of demographic variables revealed statistically significant differences in the PLR, NLR, and other variables between the two groups of patients with stages I and III gastric cancer, with a significance level of P-value < 0.05. Moreover, these findings suggest that the KNN model in this study is one of the best models for predicting the stage of GC.