Optalert pioneers new method for OSA testing

For the last decade Optalert’s focus has been driver drowsiness. We are now turning our attention to other health conditions and have developed an AI-based model that tests for obstructive sleep apnoea (OSA). The condition affects 13.2% of the world’s population and it is chronically underdiagnosed: 80% of people with OSA never undergo diagnosis or treatment. To address this challenge, we are pioneering a breakthrough OSA test. It is performed on a fully awake subject, takes minutes to complete, and the result will be instantly available!

With our innovative technology, we are laser-focused on an ambitious but achievable mission:

Introduce an OSA test that is quick, easy, and accurate.

The performance of our first version is exceptionally accurate compared to the commonly used screening tests. And it sets us up to approach a diagnostic-grade test in future iterations built on a greater volume of data.

How did we do it?

The first stage of development of the testing device involved the following two broad steps:

  1. Gather data: The data came from a meticulously controlled study involving a test group with untreated OSA and a control group without OSA. The selection criteria were very rigorous as was the test itself, which was approved by Austin Health Research Ethics Committee. All participants performed a range of vigilance tasks while having their eyelid movements recorded. The resulting data was a series of eyelid signals labelled by whether the subject had untreated OSA or no OSA.
  2. Develop a model: Our data science team specialises in working with blepharometric data involving eyelid signals. Optalert’s founder Dr. Murray Johns pioneered this field and we remain the world leaders. The resulting model had an astounding accuracy of 84%!

How does the technology work?

Blepharometry is the study of eyelid movements. Interestingly, the eyelid is controlled by two muscles in the brain. The pons stimulates the orbicularis oculi muscle, which closes the eyelid. The midbrain stimulates the levator palpebrae superioris muscle, which opens the eyelid.

Over a long period, someone with OSA gets less oxygen to their brain as they sleep. This presents as neurological degradation and impaired coordination between these two parts of the brain. The resulting model is vastly different from our Johns Drowsiness Scale (JDS) that measures impairment from drowsiness in the immediate term. Our OSA model detects the long-term impacts of the sleep disorder.

It is remarkable how much information the eyelids give us! They truly are a window into our cognitive state.

What impact will it have?

We foresee a multitude of ways our breakthrough could transform the OSA landscape, across both active and passive screening, diagnosis, and treatment.

It offers significant advantages compared to current tests:

As with all AI models, more data will drive the test’s accuracy even higher. Our goal is for it to become a diagnostic-grade test, with accuracy nearing the gold standard of polysomnography (PSG).

This technology will have a profound impact on public health outcomes. Service providers and product manufacturers across the ecosystem will benefit. Test subjects will benefit from more reliable and readily available data. People with OSA will be given a more frictionless path to diagnosis and treatment. The longer-term comorbidities that arise from untreated OSA will be inhibited, in turn reducing public health costs.

Further development is underway. Optalert is now in the process of seeking one or more partners in the OSA ecosystem to bring this technology to market. We are confident this will positively impact the health of millions of people.

Cybernetic artificial intelligence in three-dimensional logic space on quantum computing communication of abstract Plexus elements array. 3D illustration concept loop for music, logic, and meditation.

Want to learn more?

Please get in touch if you want more details on our OSA breakthrough. If you would like to collaborate with us on this journey, please contact us.