Abstract:
Deep learning algorithms in clinical imaging for the kidney in stomach computed tomography (CT) images have been constrained. Due to this, our main goal is to apply deep learning models along with a suggested training scheme to achieve precise and actual results of the division. Additionally, this group hopes to provide the community with an open-source, unedited dataset of stomach CT images for use in developing and testing deep-learning classification organizations that can segment kidneys and identify kidney disease. The proposed methodology is to detect and classify kidney disease by implementing a TensorFlow library from the computed tomography images of the human kidney using Convolutional Neural Network (CNN) architecture deep learning. In the proposed methodology after using the TensorFlow framework with VGG16 architecture, we achieved 99.13% accuracy in the case of training and 99.63% in validation respectively. Besides, we additionally looked at our model without TensorFlow VGG16, the exactness of preparing and approval precision are 99% and 99% seperately. The TensorFlow model beat the keras model with additional productive and precise outcomes.
