ML-assisted prediction of esophageal and stomach cancer

A novel study was launched for building tools for automatically predicting incident esophageal adenocarcinoma (EAC) and gastric cardia adenocarcinoma (GCA) using electronic health records. The purpose was to make screening decisions more precise and purposeful.

Accordingly, patients diagnosed with EAC (n = 8430) or GCA (n = 2965) were identified and compared with 10,256,887 controls. The Kettles Esophageal and Cardia Adenocarcinoma Prediction (K-ECAN) tool was developed and internally validated using a machine learning method of simple random sampling imputation and extreme gradient boosting.

K-ECAN was found to be well-calibrated with better discrimination vis-à-vis previously validated models. Although gastroesophageal reflux disease was strongly associated with EAC, it contributed only a small proportion of gain in information for prediction. K-ECAN is indeed an effective tool for predicting incident EAC and GCA using electronic health records but a broad-based validation is imperative for assessing the best way to implement the tool within electronic health records.

Read more: