How AI is helping doctors spot blood diseases faster
Millions of people in the UK have blood disorders. These may range from iron deficiency through to blood cancers like leukaemia. Every year, over 41,000 people find out they have blood cancer.
Usually, we need to look at blood samples under microscopes to find these diseases, which requires lots of expertise and time.
In 2023/2024 alone, the NHS East and South East London (ESEL) Pathology Partnership, which includes Barts Health NHS Trust, Homerton Healthcare NHS Foundation Trust and Lewisham and Greenwich NHS Trust, analysed a total of 98,500 blood films – that’s more than 250 per day. Each one takes 5-10 minutes to check, and that means it takes more than a whole day just to look at one day's samples.
To speed up the process, a smart computer program called BloodImage! has been created to look at blood samples and spot problems which uses AI to analyse abnormalities from blood data much faster and possibly more accurately than humans. It may even pick up patterns' humans haven’t spotted.
The team behind it includes:
Doctors from Barts Health and scientists from East and South East London Pathology Partnership
Computer experts from Barts Life Sciences which is partnership of Barts Health and Queen May University of London
Partners from universities across the UK, Europe, and Africa
Concetta Piazzese, a Research Data Engineer, built an AI model that can find cancer cells in blood pictures.
Dr. Juan Tan is testing the program at Royal London Hospital with 350 patients. So far, the model has been able to correctly identify patterns in medical images in about two out of three cases, which is a promising start.
The team want to test it with 1,000 to 3,000 more patients. If it works well, doctors will be able to:
Give patients answers much faster
Spend more time talking to patients instead of looking through microscopes
Catch diseases earlier when they're easier to treat
Faster diagnosis means better care for patients and more time for doctors to focus on what they do best - helping people get better.
How AI is helping doctors spot blood diseases faster
Millions of people in the UK have blood disorders. These may range from iron deficiency through to blood cancers like leukaemia. Every year, over 41,000 people find out they have blood cancer.
Usually, we need to look at blood samples under microscopes to find these diseases, which requires lots of expertise and time.
In 2023/2024 alone, the NHS East and South East London (ESEL) Pathology Partnership, which includes Barts Health NHS Trust, Homerton Healthcare NHS Foundation Trust and Lewisham and Greenwich NHS Trust, analysed a total of 98,500 blood films – that’s more than 250 per day. Each one takes 5-10 minutes to check, and that means it takes more than a whole day just to look at one day's samples.
To speed up the process, a smart computer program called BloodImage! has been created to look at blood samples and spot problems which uses AI to analyse abnormalities from blood data much faster and possibly more accurately than humans. It may even pick up patterns' humans haven’t spotted.
The team behind it includes:
Concetta Piazzese, a Research Data Engineer, built an AI model that can find cancer cells in blood pictures.
Dr. Juan Tan is testing the program at Royal London Hospital with 350 patients. So far, the model has been able to correctly identify patterns in medical images in about two out of three cases, which is a promising start.
The team want to test it with 1,000 to 3,000 more patients. If it works well, doctors will be able to:
Faster diagnosis means better care for patients and more time for doctors to focus on what they do best - helping people get better.
Want to know more or work with the team? Email them at datascience@bartslifesciences.org
Read more: