Intelligent Gastric Screening
Objective
The main objective of the investigation was determining a possible relation between the symptoms of Gastro-Esophageal Reflux Disease (GERD) and the result of the 24-hour pH meter due to the possibility to apply the respective symptoms in a machine learning model in order to predict the results of the study.
Background
As a background, GERD is one of the most frequent reasons for consultation in gastroenterology and is defined as a condition in which gastric content refluxes into the esophagus, causing symptoms such as heartburn, cough, among others. [1]
In addition to this, impedance pH-meter quantifies and characterizes gastroesophageal reflux, especially in patients with refractory GERD. It is important to mention that the study is complicated, expensive and, in certain cases, its application may not be necessary. [2]
Due to this, the team decided to look for a way to predict the result of the PH meter from the symptoms with the intention of avoiding applying the study in cases where it was not strictly necessary to use it, reducing costs and time.
How do we participate?
The analysis of the symptom questionnaire was performed using Machine Learning techniques with the Mathlab classification learner app. Afterward, confusion and AUROC matrices were generated to find the best model that predicted the pH-meter result. Characteristic engineering was also used to define the three-dimensional clinical space of the frequency, intensity, and duration in each of the esophageal symptoms, and the vector was then compared between the different groups for each of the symptoms. A dimensionality reduction analysis was used with the generated vectors to maximize information and variability, and a correlation matrix was used to identify groups with significant associations between symptoms.
Collaborators and Researchers
Granados Molina, Edgar Alejandro; Vela, Daniel; León Chávez, Héctor Mario; Santos Díaz, Alejandro; Guerrero Hernández, Cynthia Fernanda; Coss Adame, Enrique; González Chagolla, Alex; Vázquez González, Juan Carlos.
Participating Institutions
References
- PJ K, NJ S, MF V, SW H, E B, IM M, et al. American Gastroenterological Association Medical Position Statement on the management of gastroesophageal reflux disease. Gastroenterology [Internet]. 2008 [cited 2021 Jul 14];135(4). Available from: https://pubmed.ncbi.nlm.nih.gov/18789939/
- I H, JE R. ACG practice guidelines: esophageal reflux testing. Am J Gastroenterol [Internet]. 2007 Mar [cited 2021 Jul 14];102(3):668–85. Available from: https://pubmed.ncbi.nlm.nih.gov/17335450/