A team from the University of Huddersfield is working on the development of a secure threat intelligence sharing platform with the aim of helping to protect AI-enabled diagnostic tools from cyber attacks.
Dr Faisal Jamil, lecturer in computing within the university’s centre for autonomous and intelligence systems, is leading the development of the new software, and comments that AI models are vulnerable to cyber security breaches as traditional security methods have not been designed to prevent attacks on AI models.
The platform, having completed the market validation and proof of concept rounds in Innovate UK’s cyber security academic startup accelerator programme, uses federated learning and swarm intelligence to protect patient data. Dr Jamil also notes plans to share threat intelligence anonymously across the healthcare network to support the detection and prevention of attacks in real time, offering “collaborative advantage with improved security through using collective insights to stop attacks before they happen”.
Dr Jamil and his team are to present their work to Innovate UK at the end of this month in the hopes of securing further funding and developing the platform for commercial market readiness.
Spotlight on artificial intelligence
Yesterday we reported on the announcement of a partnership between Leeds Teaching Hospitals NHS Trust and health tech startup Newton’s Tree which will see the startup’s enterprise AI platform deployed across the trust, with the aim of supporting Leeds Teaching Hospitals to “rapidly scale” its ability to evaluate and implement artificial intelligence applications.
In June, HTN highlighted a skin cancer pilot launched in Tameside and Glossop utilising an AI platform to help triage and assess skin lesions for suspected cancer; and we shared how a Cambridge University Hospitals consultant gastroenterologist has been awarded a funding grant of up to £365,000 to explore how artificial intelligence and video can help identify gastric cancer at an earlier stage.
We also covered news from Bradford Teaching Hospitals which sees AI software utilised in its radiology department with the aim of enhancing diagnostic accuracy, efficiency and patient outcomes.
Also in Huddersfield
In May, we shared how Calderdale and Huddersfield NHS Foundation Trust has achieved HIMSS stage six validation for its use of data (Adoption Model for Analytics Maturity), with the trust demonstrating a maturity in their use of predictive analytics as well as an “expanded focus” on advanced data content and clinical support.
Earlier in the year we noted that the trust went live with the national pathology imaging co-operative digital pathology platform, described as an “high-capacity pathology image repository and exchange system”; and a new patient engagement portal was introduced with the aim of offering people more choice when booking appointments and improving two-way communication.
