Mobile Health (mHealth) Solutions for Breast Cancer Detection: A Deep Learning Approach

Breast cancer continues to be one of the leading causes of death and disability among women worldwide, and the key to increasing survival rates is finding the disease early. Most people cannot afford or do not have access to traditional diagnostic technologies, especially in countries with limited resources. This research presents a promising solution to the disturbing rates of breast cancer by designing a mobile tool that employs cutting-edge Deep Learning (DL) technology for breast cancer diagnosis, which is now a major challenge in many underdeveloped and developing countries. The application is based on DL models such as Basic CNN and MobileNet which are trained on large ultrasound datasets using data augmentation techniques to combat the limitations of the datasets. The impressive results indicated that the basic CNN model obtained 98% accuracy while the MobileNet model obtained 99% accuracy. The mobile application combines these models, offering an easy-to-use interface for real-time image analysis and initial diagnosis, especially assisting underserved or remote communities. Such innovation is very beneficial for neglected areas or populations as it helps reduce barriers to the cost and availability of cancer diagnostics. Overall, it highlights the great prospects there are in the combining of mobile Platforms and deep learning technologies in clinical practice.