Development and Validation of a Portable Device for At-Home Hydrogen and Methane Breath Testing

Authors: Riccardo Avvisati, Joshua Bates, Hannah Winter, Madeleine Ball, Lara Pocock*, Huw Davies, Rui Pinto-Lopes, Francesco Guagliardo, Martin Nash, Paul Scott, Robert McCarthy, Billy Boyle 

View the Poster PDF

Abstract

Introduction 

Hydrogen and methane breath tests can diagnose small intestinal bacterial overgrowth (SIBO) due to the connection between these gases and the activity of the gut microbiome. We have developed a portable hydrogen and methane breath analyzer device (patent-pending) to support diagnosis, longitudinal monitoring of treatment response and early detection of recurrence. The OMED Health Breath Analyzer utilizes a gas concentration predictive model developed by simulating human breath across a range of temperatures and humidities. We benchmarked the accuracy of the device to an in-clinic HMBT using the breath of volunteers. 

Methods 

941 synthetic breath exposures were included in a dataset across clinically relevant concentrations, temperatures, and humidities. The data was split into three datasets (training, validation, and test) to build a gas concentration predictive model. The model was then tested on 243 breath samples collected from 54 participants and compared to the Gastrogenius™ Breath Monitor (HMBT). Each participant gave two samples into the HMBT, and a breath sample into the OMED device. A mean of the two HMBT measurements was calculated (AVG) as the assumed ground truth of breath concentration. The results of the AVG HMBT measurements were then compared with the values from the OMED device for hydrogen and methane. 

Results 

The sensors in the OMED device and analytical model utilized were shown to assess and report hydrogen and methane levels accurately. Between the device and AVG HMBT, there was a high degree of correspondence between the concentrations detected. The errors (measured as the difference between the OMED device reading and AVG for the OMED device, and the difference between measurements 1 and 2 of HMBT for the HMBT) were comparable for both hydrogen (mean absolute error: OMED: 3.83 ppm, HMBT: 2.11 ppm) and methane (mean absolute error: OMED: 2.32 ppm, HMBT: 1.60 ppm).  

Discussion 

The OMED device is highly accurate with comparable performance to an in-clinic HMBT. The added portability for convenience and at-home measurements can benefit users by providing longitudinal data immediately available to their healthcare provider.