Artificial Intelligence Can Predict People’s Future Health Problems
Artificial intelligence can predict patients' future health problems. Scientists said this technology could help doctors in diagnosis and patient follow-up. Trained with the data of more than 811 thousand patients, the artificial intelligence tool was found to be 68 to 76 percent accurate.
According to a study, artificial intelligence can predict what kind of health problems a patient might have in the future.
Scientists suggest that this technology could be used to assist doctors in monitoring and diagnosing patients.
MORE THAN 811 THOUSAND PATIENT DATA USED
The researchers trained three different Foresight models using data from more than 811,000 patients from two NHS trusts in London and a publicly available dataset in the US.
The AI tool, known as Foresight, belongs to the same family of AI models as ChatGPT, but was trained using information from NHS electronic records.
76 PERCENT ACCURACY RATE
Based on the records, 10 possible conditions that a patient might experience in the future were considered.
When using data from NHS hospitals, Foresight was able to correctly identify the condition 68 percent and 76 percent of the time. When using US data, it was 88 percent accurate.
Zeljko Zeljko, first author of the study published in The Lancet Digital Health, said that the findings show that the AI tool has a very high accuracy in predicting patients’ health trajectories, which can be used as an important tool to guide clinical research and aid decision-making.