Gastroesophageal Reflux Disease (GERD) is a digestive tract disorder caused by the backflow of stomach acid into the esophagus, which can trigger serious complications such as esophagitis and esophageal cancer. Early detection of this disease is important, but conventional methods such as endoscopy are invasive and costly. This study designed and developed an Electronic Nose (e-nose) device based on gas sensors for early, non-invasive detection of GERD using the k-nearest neighbor (KNN) method. The system uses MQ-137 (NH), MQ-136 (H2S), and MQ- 7 (CO) gas sensors, along with a pH E201-C sensor to detect volatile compounds and the acidity level of saliva. The data collected is processed using an ESP32 microcontroller and classified using the k-nearest neighbor (KNN) algorithm. The test results on 20 samples show that the system can classify GERD and non-GERD patients with 95% accuracy, 100% precision, 90% sensitivity, and 100% specificity. These evaluation values demonstrate that the e-nose device has high detection performance, is fast, and more comfortable compared to conventional clinical methods. This innovation is expected to serve as an efficient alternative diagnostic solution and support the development of medical technology based on the Internet of Things (IoT).
Keywords: Gastroesophageal Reflux Disease (GERD), Electronic Nose, gas sensors, non-invasive detection.