Seyed Mohssen Ghafari, Richard Nichol and Richard A. George, Faethm AI Company, Australia

At the time of writing, more than seventy million people have been infected by COVID19 and more than one and a half million have died from the infection. A major challenge for health systems around the world is to supply ventilators and Intensive Care Unit (ICU) beds for those patients with the most severe symptoms of the infection. Unfortunately, during the COVID-19 pandemic, many countries face ICU bed shortages. In situations of peak-demand, healthcare providers follow predefined strategies to allocate the available ICU beds in the most efficient way. On these occasions, physicians and healthcare workers, who swore an oath to treat the ill to the best of their ability, would have to choose not to save some patients to ensure others survive. This decision puts healthcare professionals in an ethically and emotionally challenging situation in an already stressful environment. In this paper, we propose an automatic approach for managing ICU beds in hospitals to i) create the most effective ICU resource allocation, and ii) relieve physicians of having to make decisions in this regard. The experimental results demonstrate the effectiveness of our approach

COVID-19, Resource Allocation, ICU Beds, Regression