THIS MESSAGE (MATERIAL) IS CREATED AND (OR) DISTRIBUTED BY A FOREIGN MASS MEDIA PERFORMING THE FUNCTIONS OF A FOREIGN AGENT AND (OR) A RUSSIAN LEGAL ENTITY PERFORMING THE FUNCTIONS OF A FOREIGN AGENT.
Scientists have developed a neural network to predict the severity of coronavirus infection. The study can be carried out on blood taken at any stage of the coronavirus. According to TASS, the description of the development was published by the scientific journal PLoS Digital Health.
Testing has shown that this neural network copes with the assessment of the probability of death of patients better than other methods.
Scientists analyzed blood samples from five dozen patients who were hospitalized in Berlin during the first waves of the pandemic with very severe forms of COVID-19. Some of them died as a result.
Specialists collected blood samples at different stages of infection and tracked how the concentration of three hundred proteins changed over the course of the disease, as a result of deterioration or improvement in the patient’s condition. The doctors analyzed these data using a neural network created specifically for this purpose by their fellow mathematicians.
Thanks to this, scientists identified 14 key proteins, the concentrations of which most significantly influenced the outcome of the infection. Based on these biomarkers, Kurt and his colleagues prepared another neural network that determined the likelihood of a patient’s death and predicted the most likely course of infection.
As scientists hope, this will allow using the neural network they created to optimize the distribution of resources and forces of doctors in the treatment of a large number of patients.
Source: Rosbalt
Tristin is an accomplished author and journalist, known for his in-depth and engaging writing on sports. He currently works as a writer at 247 News Agency, where he has established himself as a respected voice in the sports industry.
