Cancer Detection: Detecting cancer using ultrasound by Deep Learning Techniques

Cancer is one of the major leading causes of death worldwide despite several advancements in the field of technology and medical science. The detection of cancer, especially in its early stages, is proven to be very beneficial for the recovery of the patient. However, the process often involves complex procedures and requires skilled medical professionals, making the costly cure challenging to reach remote places and poor population. In light of these challenges, there is an increasing demand for efficient and accessible diagnostic tools. This is where our proposed model comes into use. Leveraging deep learning techniques, our model basically performs the tasks below: detecting whether the image is ultrasound or not, classifying the organ present in ultrasound images, breast cancer detection, and gallbladder cancer detection. These models align with the need for accurate and fast cancer detection, which can significantly improve patient survival chances. For detecting ultrasound or not as well as organ classification, the ResNet50 model is used which identifies the ultrasound images, and accurately classifies them based on the organ they represent. This capability is crucial in determining the appropriate treatment plan for patients. The model further enhances its utility by identifying whether the breast image has any cancer or not using a VGG16 model. This feature is crucial in early detection and prevention of breast cancer. At last, the model incorporates gallbladder cancer detection using another ResNet50 model. Early detection of these diseases can significantly improve patient survival chances. The approach lowers the need for manual inspection by automating these operations, improving healthcare delivery's accuracy and efficiency. In addition, it offers a platform for ongoing education and development, advancing medical technology.