133386Automating prediction of exacerbation in respiratory disease using complex data-sets generated at home by patients. An unmet clinical need.ClosedFeasibility StudiesISCF"Incurable respiratory diseases are characterised by frequent exacerbations (flare-ups), worsened health status, rapid lung function decline caused by infections (bacterial or viral), triggered by air pollution, cold temperatures or other environmental factors. Airway inflammation and increased mucus production result in deteriorating airflow, dynamic hyperinflation and increased workload of breathing, precipitating respiratory failure. Human rhinovirus is the most common respiratory virus while the most common bacterial pathogens and _Pseudomonas aeruginosa_ _(PA)_ and _Haemophilus influenzae (HI)_, detectable days before an exacerbation is recognised. Recovery is usually 7-10 days, but can last 12 weeks before returning to baseline status. Only 20% of patients recover to pre-exacerbation levels 8 weeks later. Exacerbations cause disease progression. Severe exacerbations requiring hospital admission are associated with cardiovascular events. Once patients experience an exacerbation, they show increased susceptibility to subsequent exacerbation events with respect to severity and frequency. Early treatment improves symptoms and reduces the risk of hospitalisation or developing respiratory failure. Unfortunately, in patients with moderate to severe disease who have chronic symptoms, it is difficult to detect the onset of exacerbations until well established. Treatment comprises antibiotics and oral steroids, but overuse increases bacterial colonisation and resistance. Some patients are vulnerable to steroid side effects such as muscle wasting and osteopenia. Accurate and early prediction of exacerbations is important and is as yet, an unmet need in managing patients with Long-Term Respiratory Conditions. The long-term prognosis following an exacerbation requiring hospitalisation for COPD for example, is poor with a five-year mortality rate of 50%. Poor outcomes are associated with older age, lower body mass index, existing comorbidities (e.g. cardiovascular disease or lung cancer), a history of previous admissions for exacerbations, clinical severity, and the need for long-term oxygen therapy at discharge. In this project two UK small companies (Aseptika Ltd and Zenzium Ltd) seek to develop new tools using the latest AI techniques to help patients and their healthcare professionals automate clinical decision-making. By detecting exacerbations automatically using the patient's own self-generated data, impending exacerbations can be detected sooner which will enable the patient to receive earlier treatment, which will in turn will reduce symptoms and lower the risk of further lung damage."