According to the World Health Organization statistics, colorectal cancer is the third most common cancer worldwide, accounting for approximately 10% of all cancer cases and being the second leading cause of cancer-related deaths. Colorectal cancer is usually diagnosed when the disease has progressed significantly, making it difficult to treat and may have spread to other organs. Early and reliable diagnosis can significantly reduce the risk of death for many individuals. Artificial intelligence has made significant advancements in the field of medicine, and it can be utilized for timely disease detection. Therefore, in this research, an attempt has been made to train a convolutional neural network using histopathology data from colorectal cancer and non-cancerous tissues to distinguish between cancerous and non-cancerous colorectal tissues. Using the TensorFlow library, the convolutional neural network achieved accuracy, precision, recall, and F1-score of 98%, 97%, 99%, and 97%, respectively. Deep learning models require large amounts of data for training. Therefore, to address this issue, a generative adversarial network was employed. The generative adversarial network was trained using colorectal histopathology data, but due to limited computational power, satisfactory output was not obtained.
The dataset has been downloaded from Kaggle lung and colon cancer histopathological images.