Jiva.ai & Aevice predict asthma attacks from wearables

Jiva.ai and Aevice Health have completed a three-year programme to develop an artificial intelligence model that predicts asthma exacerbations using data from wearable monitors. The system identified more than 60% of exacerbations across prediction windows of up to 72 hours.

The project brought together Jiva.ai, a UK artificial intelligence company, and Singapore-based remote monitoring specialist Aevice Health under a programme backed by Innovate UK and Enterprise Singapore. It focused on chronic respiratory disease, including asthma, which forms part of a wider disease burden affecting more than 545 million people globally.

At the centre of the programme was Aevice Health's wearable stethoscope, AeviceMD, which is cleared by the US Food and Drug Administration. The device records continuous respiratory data, including wheeze detection, heart rate and respiratory rate, and sends the information alongside patient app inputs for analysis.

Jiva.ai applied its model-fusion technology to data gathered from the wearable device and patient reports. This allowed several streams of physiological and patient-reported information to be analysed together in a single predictive system.

The programme analysed anonymised data from 185 patients across several clinical sites. According to the companies, the resulting models were able to predict sudden asthma exacerbations within 24-hour, 48-hour and 72-hour windows.

The work also required an international framework for sharing data between the UK and Singapore. That structure was set up to support compliant analysis of anonymised patient information across borders during the three-year collaboration.

Respiratory burden

Asthma and chronic obstructive pulmonary disease are among the most common forms of chronic respiratory disease. These conditions can worsen because of infections, stress, allergens such as dust and pollen, and poor air quality.

The announcement was made on UK Clean Air Day, linking the project to broader concerns about pollution and respiratory illness. The World Health Organisation has said 99% of the global population breathes air that exceeds its guideline limits, underlining the scale of environmental exposure for people with respiratory conditions.

Predicting an exacerbation before symptoms become severe could give patients time to follow an agreed clinical plan or seek advice earlier. In practice, that might mean using medication sooner, increasing monitoring or contacting a healthcare professional before an episode leads to emergency treatment.

Remote monitoring is a growing area of interest for health systems trying to manage long-term conditions outside hospitals and clinics. Wearable devices are increasingly used to collect a continuous stream of patient data, but turning those readings into clinically useful warnings remains a challenge.

This collaboration sought to address that challenge by combining a medical monitoring device with an analytical model designed to detect patterns that might precede deterioration. The project was presented as an attempt to move from tracking respiratory status in real time to providing earlier alerts.

For Jiva.ai, the healthcare project adds to a list of industry uses for its artificial intelligence tools. Founded in 2019, the company says its systems are used by organisations including the NHS and the Dubai Health System, as well as smaller businesses.

Aevice Health focuses on asthma and chronic obstructive pulmonary disease through remote monitoring. Its wearable stethoscope and software platform are intended to allow patients to be observed over longer periods outside conventional care settings.

Dr Manish Patel outlined how the two sides divided the work during the programme. "This partnership harnessed the best of Aevice Health and Jiva.ai: they brought the novelty of capturing the data in the first place, and we brought along the expert ability to analyse it. Following successful completion of the programme, Jiva.ai and Aevice are exploring the next phase of clinical validation and regulatory development to advance the technology towards commercial deployment," said Dr Manish Patel, Chief Executive Officer and Co-founder of Jiva.ai.

Dr Rex Tan described the model as an early step rather than a finished clinical product. "Developing a predictive model for asthma exacerbations is an important step toward helping patients better understand their risk and stay engaged with their care. By identifying individuals at higher risk of deterioration, this approach can support earlier intervention, better adherence and more proactive disease management. While the initial outcomes are encouraging, further work is needed to translate this into real-world clinical practice. We look forward to continuing our collaboration with Jiva.ai to advance predictive analytics in respiratory care," said Dr Rex Tan, Chief Technology Officer of Aevice Health.