If, for example, pneumonia is suspected, the attending physician usually takes an X-ray of the chest. Researchers at Harvard University and Stralsund University of Applied Sciences are now planning to use such images to predict long-term mortality. How is this supposed to work? With the help of artificial intelligence.
For the prediction, the scientists have created an artificial neural network, which independently evaluates the image data of the X-rays and determines the probability of death. More than 55,000 images were evaluated, with around 40,000 images being used to develop the algorithm. The remaining images were then used for validation.
The final algorithm was then used to determine risk classes. It takes about half a second to classify the images and works exclusively on the basis of the image data.
According to Prof. Dr. Thomas Mayrhofer from Stralsund University, knowledge about the individualised mortality risk can be used to make informed decisions about prevention measures such as lung cancer screening.