Artificial Intelligence and Cancer Staging in Upper Gastrointestinal Malignancies

Overview

Esophageal and stomach cancers, which constitute cancers of the upper region of the digestive system, are cancers that are frequently observed and unfortunately have a low rate of cured patients. In these cases, the stage of cancer at diagnosis is very important for two reasons; First, the stage of the cancer is directly related to the survival time. Secondly, treatment is planned according to the stage. Different treatments are applied to patients at different stages. Currently, the TNM staging (Tumor, Lymph Node and Metastases) system is the accepted one worldwide. Despite many advanced technology tools used in staging (Computed Tomography, Magnetic Resonance Imaging, Endoscopic Ultrasonography), there are still difficulties in correct staging before surgery or before-after neoadjuvant therapy. Artificial intelligence techniques are increasingly used in the field of health, especially in the diagnosis and treatment of cancers. Obtaining cancer details in radiological images, which cannot be noticed by the human eye, by analyzing big data with the help of algorithms gave rise to the application area of "radiomics". It is stated that with Radiomics, there will be improvements in both the diagnosis and staging of cancers and, accordingly, in the treatment. While there are studies on the use of endoscopic methods with artificial intelligence for the early diagnosis of esophageal cancers, a limited number of studies have been conducted on stage estimation from radiological images. In particular, there are not enough studies on the investigation of changes in tumor size after chemotherapy with artificial intelligence and the estimation of staging. In this study, it was aimed to investigate the predictive efficiency of staging and the accuracy of the algorithm developed with artificial intelligence by processing tomography images in a region where esophageal cancers are endemic as a primary outcome and to evaluate the post-treatment mortality, morbidity rates and complication rates of the patients as a secondary outcome.

Conditions