it could become a crucial tool for early detection
The UK Health Security Agency (UKHSA) is exploring how AI could help detect and investigate foodborne illness outbreaks.
A new study by UKHSA experts has assessed different AI models for their ability to detect and classify text in online restaurant reviews.
The approach could one day help identify and target investigations into foodborne illness outbreaks, potentially improving public health response times and reducing the spread of illness.
Foodborne gastrointestinal (GI) illness, which causes vomiting and diarrhoea, is a significant public health burden in the UK.
Millions of people become unwell each year, but most cases are not formally diagnosed.
This makes it difficult to determine the true scale of outbreaks and their sources.
UKHSA scientists examined large language models to assess their ability to analyse thousands of online reviews for symptoms related to GI illness, such as diarrhoea, vomiting, and abdominal pain.
They also looked at reports of different foods consumed before illness.
If AI can accurately identify potential outbreaks, it could become a crucial tool for early detection and intervention.
Researchers believe AI could improve current disease surveillance by identifying cases not captured by existing systems and providing clues about potential sources of outbreaks.
Traditional methods rely on self-reporting and formal diagnoses, which means many cases go unrecorded. AI-powered monitoring could bridge this gap by analysing real-world data in real-time.
However, the study highlighted challenges that need addressing before AI can be routinely used.
Real-time data access is a key issue.
While AI can identify general food types linked to illness, pinpointing specific ingredients remains difficult. Spelling variations, slang, and misattributed illness reports were also identified as obstacles.
Additionally, privacy concerns and data-sharing restrictions could pose further challenges in implementing AI-driven surveillance.
Professor Steven Riley, Chief Data Officer at UKHSA, said:
“We are constantly looking for new and effective ways to enhance our disease surveillance.
“Using AI in this way could soon help us identify the likely source of more foodborne illness outbreaks, in combination with traditional epidemiological methods, to prevent more people becoming sick.
“Further work is needed before we adopt these methods into our routine approach to tackling foodborne illness outbreaks.”
Previous research has examined AI’s potential to analyse restaurant reviews, but UKHSA’s study expands this by using a more detailed list of terms to detect illness outbreaks.
The research involved testing various AI models to determine which could most effectively identify relevant information while minimising false positives.
Over 3,000 reviews were manually annotated by epidemiologists after being collected and filtered. Reviews containing GI-related keywords were examined for relevant symptoms.
Non-specific symptoms such as headache, fever, and respiratory issues were excluded from the analysis.
By refining AI’s ability to detect foodborne illness trends, UKHSA hopes to develop a more robust and efficient monitoring system for public health.
Loading ...