Detection of Colorectal Cancer with Vision Transformers
Mohamed Traore Mali, Emrah Hançer, Refik Samet, Zeynep Yıldırım, Nooshin Nemati · 2022 Innovations in Intelligent Systems and Applications Conference (ASYU) · 2022
Colorectal cancer (CRC) is the third most prevalent cancer in the U.S. Aside from lung cancer, it is the second leading cause of death among all kinds of cancers. According to the American Cancer Society (ACS), there are expected to be 106.180 new cases of colon cancer and 4,850 new cases of rectal cancer in the United States by 2022. The histological categorization of CRC tissue is critical to diagnosing the disease and making treatment decisions. However, it is challenging to classify CRC histological images due to the wide variety of tissue patterns. To alleviate this issue, we introduce a CRC detection methodology based on vision transformers. The performance verification of the CRC detection methodology is performed on a well-known benchmark, namely CRC-5000. According to the results, our methodology achieves significantly better diagnosis performance than a variety of recent works, one of which is also wrapped around vision transformers.