Foch Hospital has equipped itself with AI-based medical imaging software to detect lung damage caused by Covid-19. The tool, developed by Germany’s Siemens Healthineers, was successfully tested on 150 patients.
Foch Hospital (Suresnes) uses medical imaging software based on artificial intelligence (AI) to detect and quantify lung damage caused by Covid-19. Presented on May 5, 2020, “CT Pneumonia Analysis” aims to facilitate the work of radiologists in great demand during the current health crisis.
This software was tested in 150 patients at the hospital and demonstrated that pulmonary abnormalities correlated with severe symptoms of the virus.
IN COLLABORATION WITH SIEMENS HEALTHINEERS
This tool is the result of a collaboration between Foch Hospital and Siemens Healthineers. To develop this algorithm, the German company also worked with the American health group Northwell Health, the University Hospital of Basel in Switzerland and the Vancouver General Hospital in Canada. “CT Pneumonia Analysis” is currently only used for l’instant medical research. Siemens Healthineers wants to be clear on this issue: “its future availability as a product cannot be guaranteed”.
The software automatically segments the lung and pulmonary lobes. The extent of the diseased areas is automatically calculated and represented by the percentage of opacity and by a gravity score. The most affected regions appearing as the densest, and which are known to be markers of the severity of the disease, are also segmented and quantified in a few minutes.
The device is also very useful for monitoring the progression of the disease over time. “Quantifying progression or improvement under treatment, reliably and reproducible, is a valuable aid for clinicians and remains the most objective way to assess the therapeutic response. Pr Indeed, a En semi-quantitative and subjective evaluation, even carried out by experienced radiologists, suffers from variability that can penalize the quality of longitudinal follow-ups,” explained Prof Philippe Grenier.
Chest scanners are a very important source of information for the fight against Covid-19 because in some cases SARS-CoV-2 attacks the lungs of infected people. The French Society of Radiology (SFR) has decided to create a database données to list all these pictures. The aim is to boost the creation of detection algorithms, widely used in China since the beginning of the pandemic.