Innovate UK has awarded greater than £300,000 in funding to a collaboration between the NIHR HealthTech Analysis Centre in Sustainable Innovation and UK healthtech firm Sanome, to speed up the event of an AI-enabled system for the early detection of hospital-acquired infections.
The 18-month SMART grant will assist the co-design and roll-out of MEMORI, a Class IIb CE-certified software-as-a-medical-device (SaMD) platform that analyses real-time medical information to foretell an infection threat as much as seven days earlier than signs seem.
Hospital-acquired infections (HAIs) account for greater than 20 per cent of NHS mattress days annually, with analysis suggesting that between 35 and 55 per cent are preventable by means of earlier detection and intervention. Circumstances akin to pneumonia, MRSA and Clostridium difficile contribute an estimated 7.1 million extra mattress days yearly, at a value of round £2.7 billion to the NHS.
Preliminary research utilizing MEMORI’s first licensed model have already proven it outperforming the NHS-standard Nationwide Early Warning Rating (NEWS2) in detecting affected person deterioration. The brand new funding will assist the event of enhanced capabilities, together with:• Integration of further multimodal information sources akin to laboratory outcomes, prescriptions and medical notes, alongside present inputs together with important indicators and medicines• Deeper integration with Digital Affected person Document (EPR) techniques to embed insights into clinicians’ present workflows• A focused 20 per cent enchancment in predictive accuracy, extending the window for early intervention• Improved explainability and machine-learning efficiency to extend transparency and medical confidence
The upgraded MEMORI v2 platform can be validated by means of a large-scale stay deployment throughout a number of wards at Royal Devon College NHS Basis Belief, addressing what stays one of many NHS’s most persistent medical and monetary challenges.
The collaboration is supported by the NIHR Exeter Biomedical Analysis Centre and is meant to pave the way in which for broader adoption throughout the NHS. It additionally lays the foundations for a long-term partnership between Sanome and the NIHR HealthTech Analysis Centre, hosted by the Royal Devon in partnership with the College of Exeter.
Collectively, the companions purpose to construct a longitudinal, real-time view of affected person well being, initially targeted on an infection threat however designed to increase into wider preventative and personalised care pathways.
Benedikt von Thüngen, chief government and founding father of Sanome, mentioned: “Our mission is to forestall deterioration earlier than it turns into life-threatening. MEMORI exhibits how real-world NHS information, when safely unlocked, could be remodeled into actionable bedside insights utilizing multimodal AI. Working with the Exeter HealthTech Analysis Centre, with assist from Innovate UK, permits us to show each the medical and system-wide advantages of AI in one of many UK’s main NHS trusts.”
Dr Nick Kennedy, digital innovation and AI theme lead on the NIHR HealthTech Analysis Centre in Sustainable Innovation and a marketing consultant gastroenterologist on the Royal Devon, mentioned early intervention was vital. “Hospital-acquired infections stay one of many greatest threats to affected person security, notably for weak sufferers. By co-designing MEMORI, we are able to present how AI can assist clinicians, rework care and finally save lives.”
Chris Sawyer, innovation lead for digital well being at Innovate UK, added: “Supporting the secure introduction of AI into frontline NHS care is crucial for constructing a extra resilient, patient-centred well being service. This partnership is a robust instance of how innovation and medical experience can mix to sort out long-standing challenges.”
Preliminary affect information from the stay deployment is anticipated all through 2026, alongside plans for additional rollout throughout NHS trusts and healthcare organisations nationwide.
