Deep Learning Detection and Quantification of Pneumothorax in Heterogeneous Routine Chest Computed Tomography

News Research

2020-06-08

We are honored that Thomas Schlegl’s (our Machine Learning expert’s) recent article “Deep learning detection and quantification of pneumothorax in heterogeneous routine chest computed tomography” – a joint work of Medical University of Vienna and contextflow – was chosen to be featured in the April edition of the newsletter Highlights from the ESR Journals! Congrats to the entire team!

COVID-19 Update: New Features

News Research

2020-05-08

In March we announced we were working daily with our partner hospitals and clinics to outline how best we could utilize our current resources to provide additional value to radiologists fighting COVID-19. Over the last month, we have developed three additional capabilities for research use only towards this aim: 1) triaging – we flag suspected COVID-19 cases based on the presence of COVID-19-related disease patterns present in lung CTs; 2) updated SEARCH – our 3D image-based search engine now includes a mode that will automatically search for COVID-19-related patterns, providing heatmaps for the most common of these patterns (e.g. GGO) as well as percent of total lung volume affected; and 3) PDF reporting – we will now automatically generate a PDF report for suspected COVID-19 patients. Thanks to our partner integrators, these developments will be incorporated into normal clinical workflow to as many hospitals as possible. In fact, some are already installed and running. These accomplishments were made possible from data contributed by both our existing partners as well as new ones in Italy and Spain, to all of whom we say THANK YOU. Your courage and determination under pressure keeps us focused and motivated, and we’re truly grateful for your inspiration. Stay healthy friends.