Adaptive Jellyfish Search Optimization Trained Deep Learning for Breast Cancer Classification Using Histopathological Images
P. Vijaya · Mapana Journal of Sciences · 2023
One of the most common cancers identified worldwide among women is breast cancer, which has become the foremost reason for death. Due to complexity of breast tissues, it is essential to accurately detect and classify breast cancer in medical imaging. In this paper, an optimization-based deep learning technique is created to detect breast cancer using histopathology pictures. The pre-processed images are sent to the segmentation procedure, where the Fuzzy Local Information C-Means (FLICM) technique is used to segment the blood cells. The segmentation process is followed by feature extraction, where various features, such as area, shape, diameter, and Speeded-Up Robust Features (SURF), are extracted from histopathological images. Lastly, Deep Q Network (DQN) trained using the developed Adaptive Jelly Fish Search Optimization (AJSO) algorithm is used to identify and classify benign and malignant breast cancer. The devised approach has good accuracy, TPR, and TNR values of 91.1%, 92.1%, and 92%, correspondingly.