Artificial intelligence has learned to see the harm of hypertension earlier than a tonometer

British scientists have developed an algorithm called HyperScore that can detect hidden internal organ damage caused by hypertension in advance. A team from Oxford trained the artificial intelligence on medical data and MRIs of 32,000 people. The system analyzes the condition of the heart, brain, kidneys, liver, and blood vessels, and then determines how blood pressure levels are already affecting quality of life according to a specific scenario.

The main finding is that hypertension affects different organs in different people, and dangerous changes can appear even with mildly elevated blood pressure. HyperScore is capable of calculating the personal risk of stroke, heart failure, and kidney failure long before severe consequences occur. The most sensitive signal of hidden damage turned out to be changes in brain structures, which can be seen on an MRI.