In response to the COVID-19 pandemic, CHIMiRA mathematicians established an international, interdisciplinary, consortium comprising expertise in machine learning, public health, and medicine. They used machine learning to develop a tool based on anomaly detection which could identify the outbreak of COVID-19 in Cambridge just using full blood count data. In 2021, BloodCounts won the prestigious inaugural £1m Trinity Challenge Prize – an international award for demonstrating how “data and analytics [could] be used to better identify, respond to, and recover from global health emergencies”.

One key advantage of the BloodCounts! approach, is that it is “pathogen agnostic” and it doesn’t require prior knowledge of a specific disease pathogen. The team soon realised the tool’s potential as an early-warning system for other infectious disease outbreaks and the earlier diagnosis of non-infectious conditions. This could be particularly valuable for patients who are not exhibiting symptoms.

Current application areas being explored include better prediction of bleed vs clotting strokes, studying maternal health, improving early detection of lung and kidney cancer; pre-empting malaria and dengue fever endemics; and identifying potential new, unknown, pandemics. Developing a test for non-anaemic iron deficiency is underway, and has recently been awarded funding from the Medical Research Council and the Gates Foundation for algorithm deployment in pregnancy care in the UK, Netherlands, and The Gambia.

To make BloodCounts! as effective as possible, it is important that the tool can access blood test data from diverse patient populations from across the globe. This raises the challenges of data privacy and security. The team has developed and deployed an advanced machine learning technique, known as Federated Learning, that currently allows five hospitals in three continents, to jointly train models without sharing their data. This is scaling to more sites across the world over the coming five years with support from the Singaporean National Research Foundation.