Using AI for syndromic surveillance - 'are we nearly there yet?'
PHE ePoster Library. Morbey R. 09/12/19; 274488; 49
Dr. Roger Morbey
Dr. Roger Morbey
Login now to access Regular content available to all registered users.

You may also access this content "anytime, anywhere" with the Free MULTILEARNING App for iOS and Android
Abstract
Rate & Comment (0)
Abstract Syndromic surveillance routinely uses ‘big data' to provide early warning and real-time situational awareness. This daily surveillance involves thousands of analyses and decisions on what requires public health action and is therefore an ideal application for artificial intelligence (AI). AI aims to automate decision-making by using computers to replicate the way people think. Here we describe a number of AI techniques that have been applied to PHE's syndromic surveillance service.An important branch of AI is ‘machine learning', which involves creating algorithms that can learn from training data and automate decision-making. When examples of past decisions are available this learning is ‘supervised', otherwise ‘unsupervised' learning can uncover structure within data.We use unsupervised learning in our anomaly detection algorithms to monitor 13,000+ daily syndromic signals, e.g. GP consultations for diarrhoea in London. Unusual activity is identified and prioritised to provide epidemiologists with a manageable number of alarms. These algorithms are automatically updated with new training data; however, we adjust prioritisation rules manually, based on algorithm performance.To aid manual risk assessment of anomalies, AI (supervised learning) is used to provide guidance scores indicating how unusual are the excess counts for each alarm.During 2018, an AI Naïve Bayes classifier was trialled, to replicate the manual decision-making process during risk assessment.AI is already a necessary part of routine syndromic surveillance, with potential to extend these AI systems and make them more adaptive. External funding details
Abstract Syndromic surveillance routinely uses ‘big data' to provide early warning and real-time situational awareness. This daily surveillance involves thousands of analyses and decisions on what requires public health action and is therefore an ideal application for artificial intelligence (AI). AI aims to automate decision-making by using computers to replicate the way people think. Here we describe a number of AI techniques that have been applied to PHE's syndromic surveillance service.An important branch of AI is ‘machine learning', which involves creating algorithms that can learn from training data and automate decision-making. When examples of past decisions are available this learning is ‘supervised', otherwise ‘unsupervised' learning can uncover structure within data.We use unsupervised learning in our anomaly detection algorithms to monitor 13,000+ daily syndromic signals, e.g. GP consultations for diarrhoea in London. Unusual activity is identified and prioritised to provide epidemiologists with a manageable number of alarms. These algorithms are automatically updated with new training data; however, we adjust prioritisation rules manually, based on algorithm performance.To aid manual risk assessment of anomalies, AI (supervised learning) is used to provide guidance scores indicating how unusual are the excess counts for each alarm.During 2018, an AI Naïve Bayes classifier was trialled, to replicate the manual decision-making process during risk assessment.AI is already a necessary part of routine syndromic surveillance, with potential to extend these AI systems and make them more adaptive. External funding details
    This eLearning portal is powered by:
    This eLearning portal is powered by MULTIEPORTAL
Anonymous User Privacy Preferences

Strictly Necessary Cookies (Always Active)

MULTILEARNING platforms and tools hereinafter referred as “MLG SOFTWARE” are provided to you as pure educational platforms/services requiring cookies to operate. In the case of the MLG SOFTWARE, cookies are essential for the Platform to function properly for the provision of education. If these cookies are disabled, a large subset of the functionality provided by the Platform will either be unavailable or cease to work as expected. The MLG SOFTWARE do not capture non-essential activities such as menu items and listings you click on or pages viewed.


Performance Cookies

Performance cookies are used to analyse how visitors use a website in order to provide a better user experience.


Save Settings