Scientists have made a step forward in diagnosing complex diseases such as long Covid and myalgic encephalomyelitis (ME), more commonly known as chronic fatigue syndrome. To do this, they used a new artificial intelligence platform that demonstrates up to 90% accuracy in detecting ME using conventional laboratory tests. This is reported by the Financial Times, writes UNN.
Details
The study, published in Nature Medicine, is based on data from 249 people, 153 of whom were ME patients. The team analyzed biological changes in gut bacteria, immune responses, and metabolism, finding links between the microbiome, immune system, and chemicals involved in maintaining vital functions.
"Our goal is to create a detailed map of how the immune system interacts with gut bacteria and the chemicals they produce," said Julia Oh, a microbiologist at Duke University. "By connecting these dots, we can begin to understand what is driving the disease and pave the way for truly precise medicine that has long been out of reach."
Researchers found that ME patients have impaired interaction between the microbiome, metabolites, and the immune system, as well as reduced levels of butyrate — a substance that plays an important role in gut function. Symptoms of ME include prolonged fatigue after exertion, sleep disturbances, difficulty concentrating and memory — similar to the manifestations of long Covid.
According to Derya Unutmaz, a professor of immunology at Jackson, the lack of clear laboratory markers has led some doctors to doubt that ME is a real disease.
The problem may not be in one broken component, but in a disrupted connection between systems
At the same time, researchers warn: despite success in identifying physiological abnormalities, there are still many unresolved questions, particularly regarding the causes and effective treatment of ME.
"Patients are often diagnosed some time after the onset of the disease, which means that the causes are very difficult to determine at the molecular level," said Daniel Davis, professor of immunology at Imperial College London. "The search for effective treatments is ongoing, but the basic knowledge contained in this analysis can be used for many years to come."
Not all scientists believe that the research provides definitive answers. Some emphasize the limitations of the sample and the diversity of conditions among patients.
"At best, these (studies) are small incremental steps that are not reproducible," said Alan Carson, professor of neuropsychiatry at the University of Edinburgh. "We are still very far from understanding the biology of ME."
According to the WHO, about 6% of those who have had Covid-19 subsequently face long-term consequences. And artificial intelligence can become one of the keys to solving this global medical problem.
AI-powered microscope predicts brain disease formation - study24.07.25, 16:44 • [views_3789]