Infra-red test to help target life-saving treatment for premature babies

A life-threatening condition that affects half of all premature babies born before 32 weeks could be detected by a diagnostic test being developed in the UK.

Neonatal respiratory distress syndrome (NRDS) is the most common cause of death in premature babies, and is caused by deficiency of surfactant in under-developed lungs.

The condition, which also affects one per cent of babies born at full term, prevents the lungs absorbing enough oxygen, and can lead to complications such as chronic lung changes and bleeding in the brain.

The syndrome can be treated with administration of surfactant. However, not all premature babies need the treatment, which can damage their sensitive airways if given incorrectly.

Now a highly sensitive test that uses a mid-infrared scanner and machine learning to detect surfactant levels in babies’ gastric fluid has been developed to predict those that will go on to develop NRDS.

The tool, being developed by Stevenage-based SIME Diagnostics, takes approximately five minutes to produce a result, according to chief executive Povl Verder. It was recently published in the journal ACTA Paediatrica.

“The problem is we don’t know which babies have immature lungs at birth,” said Verder. “They may seem fine at birth, but then if they are left for a couple of hours they can suddenly become ill with respiratory distress syndrome, where they are not able to absorb enough oxygen,” he said.

The test is non-invasive, using a sample of waste gastric fluid that doctors routinely remove by suction from premature babies in order to clear their airways, Verder said.

A mid-infrared light source is targeted at the sample, causing all of the molecules within it to vibrate. Each type of molecule will vibrate at a different frequency.

“The light source will generate a biochemical signature of everything in the sample,” Verder said.

A cloud-based machine learning system is then used to analyse this signature, to identify the presence, or lack, of surfactant.

The test can be carried out with minimal training, using a tablet computer, Verder said.

The company is now carrying out further engineering work to optimise the tool for use as a bedside test within neonatal intensive care units.

In the future, the test could also be used to diagnose other conditions, such as diabetes.

Microsoft is providing the system with access to the Cloud.