Medical data mining is an important research direction in the data mining field. It has long remained a research hotspot in the computer and medical fields. Data mining algorithms usually assume that the sample distribution of data is balanced and the misclassification cost is equal. However, due to the characteristics of medical research, there are severe sample imbalances in the clinical diagnostic data collected, which leads to poor classification performance by data mining algorithms. For addressing the sample-imbalance problem in clinical diagnosis data, this paper proposes a sample filtering clustering oversampling algorithm (SFC-SMOTE). First, the k-means algorithm is used to group the clinical diagnosis data so that the similarity of the samples within the group is high while the similarity of the samples among the groups is low. Next, the sample filtering strategy is used to identify and filter the “noisy samples” in the minority and majority sets. The filtered samples of the two classes are then merged according to their original class. Finally, the dynamic rate strategy is set according to the original two-class sample numbers, and the minority and majority samples are dynamically synthesized to restore the original structure of the data. Experimental results show that three classification algorithms have the best classification effect on the dataset sampled by the SFC-SMOTE algorithm. Among them, the average Sensitivity, Specificity, and MCC of Random Forest are 92.48%, 92.07%, and 84.57%, respectively, which are significantly better than those of the existing algorithms. In addition, the Wilcoxon and Friedman test results show that the SFC-SMOTE algorithm is considerably better than the existing algorithms. The classification performance of the three algorithms has been significantly improved after the application of sampling algorithms. In addition, the oversampling algorithm is substantially better than the clustering oversampling algorithm. Importantly, regardless of the algorithm used, the classification performance of the dataset sampled by the SFC-SMOTE algorithm is superior.