Investigation of Convolutional Neural Networks for the Identification of Breast Cancer

Abstract

Early diagnosis is crucial for the identification of breast cancer, and this research investigates the use of convolutional neural networks (CNNs) to analyze ultrasound pictures and distinguish between benign, malignant, and normal breast tissue. The process starts with ultrasound image capture, then moves on to segmentation to identify areas of interest and image enhancing methods to increase quality. Using a pre-trained ResNet101 architecture, the Convolutional Neural Network (CNN) model is optimized for feature extraction and classification. 1578 tagged ultrasound pictures classified into benign, malignant, and normal classifications make up the Breast Ultrasound Pictures Dataset (BUSI), which is used to train and assess the model. The model demonstrated its capacity to reliably categorize breast tissue types with a validation accuracy of 99%, indicating exceptional performance. Precision, recall, F1-score, and confusion matrix are important assessment metrics that demonstrate the model's ability to distinguish between benign, malignant, and normal instances with little misclassification. The findings also show that the model can generalize well to fresh data, which qualifies it for use in clinical settings.