Facebook and NYU  are using AI-based models to forecast the condition of covid-19 patients in advance.  NYU plans to make the model freely available for researchers and health care providers. 
Currently, the chest-Xray of covid patients are evaluated by trained Radiologists, a costly and time-consuming process. However, the AI-based models can expedite the process at a much lower cost. In this case, the machine learning model is trained using the training set of Covid-19 patients' chest X-rays.  Once the model is trained, then chest X-rays of the patient take over several days are applied to the model to predict the condition of the patients.  The NYU study found that the model is 80% accurate when a sequence of x-ray spanning over several days is used instead of a single X-ray, which is only 70% accurate.