Machine Learning Techniques for Renal Carcinoma Detection: A Review

The sixth most common malignant disease is renal cell carcinoma (RCC), which accounts for almost 90% of all cases. The probability of recovery can be greatly increased by early detection of kidney tumors. Imaging techniques are the most sought out non-invasive diagnostic techniques in recent times. In this review, the current method of image processing involves rapid identification of the tumor from CT scans by a combination of pre-processing, segmentation, feature extraction and classification methods. Machine learning (ML) techniques such as support vector machines (SVM), random forests (RF), artificial neural networks (ANN), convolutional neural networks (CNN), and Deep learning models can be used to analyze abdominal CT images for renal cancer detection. On the other hand, a probabilistic neural network (PNN) can also be used as a feedforward neural network that uses a probabilistic methodology by accurately classifying the class probability estimates. Various types of RCCs were defined in the histological study and diagnosed using various imaging methods such as ultrasound, MRI, CT, PET and angiography in the radiological part. Machine learning algorithms are used to identify tumors, extract relevant data from medical images, and help interpret imaging results. When these algorithms are integrated with medical imaging data such as computed tomography (CT) images, renal cancer detection models become more accurate. By being trained on labeled data sets, these algorithms can identify renal cancer. A detailed review has been done and after comparing the accuracy of several algorithms, CNN is considered to be the most suitable algorithm for identification of RCC