ABSTRACT
In this research, we aim to perform diagnosis and prognosis for advanced prostate cancer patients by using machine learning methods based on CT & MRI scans of 600 patients. Measurements Four models are applied: naive bayes (NB), k-nearest neighbors (KNN), recurrent neural networks (RNN) and convolutional neural networks (CNN). CNNs perform Image Analysis with an outstanding accuracy level of 98.99%, RNNs capture Temporal Dependencies with an impressive 96.5%accuracy. Upon training, NB and KNN achieve 93.5% and 92.7% accuracy respectively. These results stress the importance of choosing the right model for a given dataset and clinical context, which might lead to better patient outcomes and care.
