This is a system for evaluating patients suspected of having COVID-19 from their chest x-rays and other basic patient information such as age or number of days with symptoms. The assessment and indicators offered by this tool can help physicians to have a more complete picture for decision making regarding patients with mild symptoms.
The software gives physicians an immediate analysis using Artificial Intelligence and Deep Learning techniques. For training the model information from different databases has been used; this training data set will be updated with information provided by several hospitals in Madrid, Spain.
The team has been the winner of the ‘Health and Social’ category and Master Winner of the Hackathon #CodeTheCurve organized by UNESCO, IBM and SAP during April 2020. This project has also been one of the top selected projects at the Spanish National Hackathon #VenceAlVirus promoted by the Community of Madrid on 4 and 5 April 2020.
This tool currently is not for diagnostic use. It is not a clinical or radiological decision-making tool. To ensure maximum privacy, all information is handled locally. The Artificial Intelligence model is downloaded as a client into the the browser of the local computer and performs the analysis without sharing information with external sources or machines.
This tool seeks to complement clinical decision-making. We hope it will be useful and can help you in your work.