Researchers listen to their guts

Inspired by acoustic sensing technology devised to pick up the munching sounds of termites, Australian scientists have created a belt to record the internal gut rumblings that would normally cause embarrassment at dinner parties.

But the belt – the brainchild of Nobel Laureate Barry Marshall – is designed to offer a safe, non-invasive tool for monitoring gut noises to help doctors diagnose intestinal disorders, starting with irritable bowel syndrome (IBS).

A true collaborative effort across multiple sectors, the Noisy Guts project was founded by a multidisciplinary team of researchers led by Mary Webberley from the University of Western Australia, in collaboration with the ON science and technology accelerator program, part of Australia’s peak science research body, the CSIRO.

Originally, Webberley wanted to study medicine, but it wasn’t an option because she faints at the sight of blood. So she applied her biological skills to becoming an expert in ladybirds and their coevolution with parasites.

Enter Marshall, the gastroenterologist who infected himself with Helicobacter pylori to prove that ulcers are caused by bacteria, not stress – a stunt that eventually resulted, in 2005, in a Nobel.

With his characteristic ability to think outside the square, Marshall heard a pitch about termite sensing technology and thought it might have use in the field of gut medicine.

Not only did the prospect open opportunities for Webberley to do research with “real-world impact”, she says, but it prompted her to follow his lead in self-experimentation.

First, the researchers combed the literature and found data suggesting a correlation between gut sounds – called ‘borborygmi’, derived from an ancient Greek name used by Hippocrates – and gut disorders.

From that, they developed a “minimum viable product” – a technical prototype using an off-the-shelf recording unit and sensors incorporating basic electronics components from local retailers.

The team members have all worn the belt on multiple occasions while moving through different iterations of the sensors. “So we’re all familiar with each other’s gurgling and rumbling and grumbling sounds,” Webberley says.

“My stomach’s particularly noisy so mine’s the test stomach that we listen to if we want to work out if sensors are overloaded by particularly noisy stomachs.”

Originally starting out with around 300 sounds, artificial intelligence-driven machine learning helped them develop 26 features to optimise the model and the software that underpins the belt.

A validation study with 98 participants – half with a clear-cut diagnosis of IBS and half with healthy guts – was able to diagnose IBS with 87% accuracy, after filtering out other physiological noises like heart beat and breathing.

Diagnosis involves wearing the belt for two hours and 40 minutes, broken up into two sittings – before and after meals – to identify noises generated by the juices and gases of digestion.

To help patients monitor symptoms after different treatments, they would wear it for 20 minutes twice a day after meals.

The next step was to join CSIRO’s ON Accelerate program to learn how to commercialise the product and take it to the real world.

The team has now secured funding to build a working prototype of the belt that is comfortable to wear and has robust sensors. It will then need further funding and larger validation studies.

The team recently won the Perth regional final of a major pitching competition, and has been shortlisted for the 2018 WA Innovator of the Year award. Webberley has been awarded the ON program’s Breakout Female Scientist Award.

IBS is a debilitating disorder normally diagnosed using a process of elimination. Characterised by abdominal pain, bloating, diarrhoea and/or constipation, it affects 11% (800 million) people worldwide.

By enabling doctors to screen, diagnose and monitor the condition, the researchers hope that the belt will bypass costly, invasive tests like colonoscopies, speeding up diagnosis for patients and enabling easier monitoring of symptoms.