Differences in coughs from people with COVID-19 To a healthy person is something that cannot be perceived and heard with the human ear. But a research team from the Massachusetts Institute of Technology or MIT In the usa It has developed a machine learning algorithm that applies advanced statistics to learn, predict, and pattern data. It is artificial intelligence or AI (Artificial Intelligence-AI) that joker123 hears, perceives and differentiates people infected with the coronavirus New species
The team relied on developments from a similar technology to detect signs of Alzheimer's disease. It uses thousands of cough and speech samples to train AI to detect people infected with COVID-19. Accurate to 98.5%, suitable for people who have been tested for the coronavirus and the result is positive. But does not show any symptoms But AI detects every time The team explains that the ResNet50 neural network is an algorithm designed to work similarly to the human brain. They were trained to distinguish sounds with different levels of vocal cord strength. In addition, two other neural networks were used to detect the emotional state of speech: frustration, happiness, total peace. To detect lung changes Respiratory function from coughing
Combining these 3 neural networks The algorithm detects muscle deterioration, giving researchers an artificial intelligence model that can find examples of Alzheimer's disease. It can also be adapted to a diagnosis of COVID-19.
