AI could help NHS clinicians diagnose childhood sleep apnoea
AI software designed to help diagnose sleep apnoea in children has achieved high accuracy in an NHS trial, with the findings independently reviewed by clinical specialists.
Seluna’s AI platform turns hours of raw neurological
data into clinical-grade insights, fast.
With a modular architecture for multi-disease
indications for all ages, our beachhead is sleep
disorders.
THE BEACHHEAD – SLEEP
5% of children & 20% of adults have sleep disorders. Untreated, it causes long-term neurological damage.
Sleep studies produce up to 144 million data points. Manual data analysis by expert sleep technologists takes up to 4 hours, particularly in pediatrics.
Limited sleep technologists means sleep labs are operating at capacity, drowning in data as waiting lists grow.
THE SELUNA SOLUTION
Hardware agnostic, ingesting data from existing polysomnography & HSAT systems.
AI & ML algorithms identify & highlight pathological regions of sleep disordered breathing.
Highest data complexity first. AI models that works on kids works on adults.
Estimated reduction in sleep study analysis & reporting time
Staff hours released per 1,000 studies
Novel digital biomarkers to drive personalized therapeutics
Clinically validated for sleep disordered breathing as part of a 500-patient UK NHS study.
THE ROADMAP
As the most data-complex indication in neurology, sleep disorders is a deliberate entry point.
AI and ML models validated here can be rapidly extended for lower-complexity indications, de-risking multi-indication neurology expansion.
Sleep
Neural-only EEG
Neuromuscular
Neural Biomarkers
Simplifying and streamlining the diagnostic pathway for paediatric sleep disorders.
Our machine learning algorithms retain diagnostic power even in the event of signal loss and noise, therefore reducing the number of inconclusive sleep studies.
Our software pipeline automatically adjusts to account for variations in data and diagnostics due to the presence of comorbidities, ensuring accurate and reliable results.
By generating detailed, patient-specific clinical reports with objective insights and full data visualization, our software supports clinicians to make a diagnosis with reduced human error.
Bias in healthcare data is real, and we treat it as a first-order problem. We actively test our models for it and engineer it out, so the results clinicians rely on are fair, transparent, and built for the real world.
At Seluna, we value open communication. Here are answers to some of the most frequently asked questions about our technology and its impact.
No, not at all. Our goal at Seluna is to support clinicians, not replace them. By building tools that can quickly filter out any redundant information, we can give clinicians a clear picture of the information that matters, allowing them to make a diagnosis faster.
Healthcare inequalities are a widespread issue. For instance, someone in central London likely has better access to healthcare than someone in rural Scotland, and a single woman without children has more time for GP visits than a mother of three. Machine learning models learn by example and are typically trained on real-world datasets. A model trained on this data might assume people in rural Scotland need fewer healthcare resources than those living in cities, masking the fundamental issue of limited access. This behaviour has severe consequences if not analysed, regulated, or monitored.
Artificial intelligence is excellent at learning how to solve complex problems extremely quickly. Artificial intelligence algorithms can learn from real-world data to match the diagnostic capabilities of experts in a fraction of the time, freeing up professionals to focus on other critical tasks.
This is a critical question that researchers are still answering today. At Seluna, we conduct extensive analysis to understand exactly how our algorithms behave. Identifying and correcting flaws early ensures our algorithms are fair, transparent, and trustworthy before they reach end-users.
Children are generally overlooked and underrepresented in the medical device industry due to a relatively small market size in comparison to adults. However, they have the longest to live and the most to gain from early diagnosis and treatment. By treating children early, we can reduce long-term strain on healthcare providers through the prevention of chronic health conditions. This makes children the prime target for preventative healthcare.
AI software designed to help diagnose sleep apnoea in children has achieved high accuracy in an NHS trial, with the findings independently reviewed by clinical specialists.
Seluna secures nearly £650K funding to advance machine learning technology that could revolutionise diagnosis and management of childhood sleep apnoea
Yola will be driving our technological innovation, developing the cutting-edge solutions that have the potential to transform healthcare delivery.