Advancements in Medical Imaging: A Deep Learning Approach for Kidney Disease Classification

Elderly age increases the risk of developing kidney tumors, a kind of cancer. As a result, It is critical to exercise caution and undergo diagnostic testing in the latter stages of life. Medical imaging and deep learning techniques are becoming increasingly appealing in this regard. It becomes essential to develop deep learning models that can help clinicians accurately identify tumors through segmentation. The soft tissues of organs, among them the kidneys and prostate, are challenging to segment, making it difficult to develop viable systems. In circumstances where detection is challenging, deep learning models are typically utilized. This paper presents a model incorporating the most advantageous elements of many existing deep learning algorithms, implementing four different deep learning models. Our evaluated models illustrate a more successful system that incorporates previously unapplied advancements in the encoder and decoder stages. Our suggested methodology might benefit most medical professionals, especially those who specialize in kidney as well as renal tumor diagnosis. The suggested model outperforms previous imaging models in detection and, because of its flexible structure and adaptability, may be simply implemented into any system. We suggested four models, MobileNetV2, ResNet50, VGG16, and VGG19, which demonstrated the training accuracy of 99.56%, 98.53%, 98.21%, and 98.41%, respectively, and validation accuracy of 99.04%, 98.07%, 98.15%, and 98.47%, respectively, for kidney and tumor identification, indicating that they could be a useful tool for studying soft tissue organ illnesses.