Abstract:
Breast cancer is still a major global health challenge that requires precise diagnosis techniques in order to plan
appropriate therapy. Because traditional methods are frequently imprecise, research into machine learning algorithms is necessary to
increase detection rates. Breast cancer affects women worldwide and has an increasing recurrence rate, major complications, and rates
of death. Benign and malignant are the two main types for cancers. This work looks into the detection and classification of breast
cancer in ultrasound images using the NASNet model, a convolutional neural network well-known for its image analysis powers.
Specifically, the NASNet Mobile model is trained on ultrasound images of breast cancer by using annotated data for supervised
learning. The model delivers outstanding performance measures, such as an Accuracy of 94.6%, Precision of 97%, and F1-score of
96%, through intensive training and validation. Its 97% Recall rate demonstrates how well it works to reduce false negatives, which is
important for early detection. Enhancing diagnostic accuracy and improving patient outcomes, the clinical practice of healthcare
providers can be greatly benefitted by the successful use of the NASNet Mobile model in breast cancer ultrasound imaging. Potential
directions for future research could include enhancing the model for wider clinical application and launching a new phase of precision
medicine in the treatment of breast cancer.
