UCLH clinicians and UCL computer scientists have got together at a sell-out UCLH/UCL workshop event to discuss challenges in healthcare ripe for the application of artificial intelligence (AI) and data science.
The event was aimed at matching up clinicians with computer scientists to find ways of using AI to improve diagnosis, speed up treatment or personalise care.
Staff from across UCLH and UCL attended the workshop, and consultant haematologist Dr Wai Keong Wong was among the clinicians to put forward suggestions for projects. His idea involved using AI and machine learning to derive additional information from the raw data produced by full blood count analysers. Dr Wong said:
‘At the moment, it takes time for both scientists and clinicians to review the results of a full blood count, prepare and examine a blood film under a microscope to decide whether the patient’s condition is life-threatening, serious but not life-threatening, or benign.
‘If we can use AI to tell us immediately how serious a patient’s condition is, we could begin treatment or further investigation where necessary without delay. Lives would be saved and outcomes for all patients would be better.
‘We have the data we need for an AI to work on this problem. Over 400,000 full blood counts are done at UCLH every year on patients. There is also a dataset of 40000 samples from healthy blood donors. A system using AI could look for patterns in this data, and learn to recognise patterns in the raw data used to derive a full blood count are saying about a patient’s condition.’
Other proposals included using AI to improve the quality of diagnostic images, and to predict common prescriptions. UCLH and UCL are now planning how the ideas presented can be taken forward.