Scientists unveiled Delphi-2M, an AI model predicting over 1,000 diseases years ahead, though experts caution it requires further testing before clinical use.
Scientists have developed an artificial intelligence model capable of predicting medical diagnoses years before symptoms appear, raising hopes for earlier intervention in healthcare.
The Delphi-2M system, described Wednesday in Nature, uses the same “transformer” technology behind consumer chatbots like ChatGPT. Trained on Britain’s UK Biobank — containing records of about 500,000 participants — the model “predicts the rates of more than 1,000 diseases” years in advance, researchers said.
“Understanding a sequence of medical diagnoses is a bit like learning the grammar in a text,” explained Moritz Gerstung of the German Cancer Research Center. Delphi-2M, he added, “learns the patterns in healthcare data, preceding diagnoses, in which combinations they occur and in which succession,” enabling “very meaningful and health-relevant predictions.”
Charts presented by Gerstung suggested the tool can identify patients at significantly higher or lower risk of heart attacks compared with traditional assessments. Verification against Denmark’s health database of nearly two million people confirmed its predictive accuracy.
However, scientists stressed Delphi-2M is not yet ready for clinical use. “This is still a long way from improved healthcare,” said Peter Bannister, a fellow at Britain’s Institution of Engineering and Technology, noting dataset biases.
Still, experts said such models could one day guide preventative medicine and resource allocation. “It looks to be a significant step towards scalable, interpretable and most importantly ethically responsible predictive modelling,” said King’s College London professor Gustavo Sudre.
THIS WAS ORIGINALLY PUBLISHED ON VANGUARD