Malignant Tumor Classification Using Hybrid CNN-SVM Models in Medical Imaging

Automatic classification of malignant tumors has become particularly important in medical imaging because early detection and accurate diagnosis can dramatically improve patient outcomes. These days, deep learning techniques have demonstrated an out-sized success in numerous fields of medical imaging, such as tumor classification. In this paper, we present a hybrid CNN-SVM model used for classifying malignant tumors in medical imaging. In this paper, we take the fugitive approach by using CNN to segment tumor regions, and SVM is used to classify cancer as Benign or Malignant. For categorization, CNN is employed to eliminate the high-level features from the medical pictures sustained by that SVM. Our study is implemented as a hybrid model that learns from an expansive collection of medical images with benign and malignant tumors. Finally, to evaluate the performance of our model, we compared it with referenced models (traditional CNN and SVM) on the same dataset. The results indicated that the hybrid model CNN-SVM is superior in terms of higher accuracy and fewer error rates. Additional experiments on a new dataset demonstrate the model's generalization capability, which can be beneficial in practical applications.