Request a Quote

Machine-made data models improve predictions for premature death

April 26, 2019

Data entry on computer keyboard


Researchers analyze a lot of data about specific diseases to discover new biomarkers and better predict treatment outcomes.

University of Nottingham UK researchers, however, have taken a wide-angle approach to data and disease, enabling computers to essentially teach themselves to better predict early death from chronic disease in a large middle-aged population.

The team started with health data from just over half a million people aged between 40 and 69 who were recruited to the UK Biobank between 2006 and 2010, and who were followed up with until 2016.

Computers built models that accounted for a wide range of demographic, biometric, clinical and lifestyle factors for each individual assessed, including their daily dietary consumption of fruit, vegetables and meat, said Assistant Professor of Epidemiology and Data Science, Dr. Stephen Weng, the study leader.

Learn More About Biospecimen Data

‘More accurate than the standard models’

“We mapped the resulting predictions to mortality data from the cohort, using Office of National Statistics death records, the UK cancer registry and ‘hospital episodes’ statistics,” he said. “We found machine-learned algorithms were significantly more accurate in predicting death than the standard prediction models developed by a human expert.”

The researchers predict that artificial intelligence will play a role in personalized medicine, tailoring risk management to individual patients. Further research requires verifying and validating these AI algorithms in other population groups and exploring ways to implement these systems into routine healthcare.

Human biospecimens = data

Large volumes of medical data like these begin with individual patients. Participants in the UK Biobank project provided biofluids such as blood, urine and saliva samples for analysis, as well as detailed information about themselves. Every human biospecimen sample yields data that can be studied, analyzed, adapted and leveraged to spawn new discoveries.

Increasingly, it appears, computers will be powering the insights.

Learn about the iSpecimen Marketplace where you can browse millions of richly annotated, de-identified human tissue and biofluid biospecimens, in addition to hematopoietic and immune cell products. You can join for free and creating a login is easy. Request a quote or custom collection today.


Request a Quote