Exploiting Artificial Intelligence to study radiographic images of patients with COVID-19 (collected and made available between hospitals and public and private research institutes), can be a diagnostic and predictive means of the various clinical developments in the fight against the Coronavirus.
The Italian Diagnostic Center (CDI) has created an information platform: the AIforCOVID Imaging Archive, in order to promote the sharing of radiological and clinical data on COVID-19 with both the national (i.e. Italian) and international scientific community.
The ultimate goal of this initiative is to foster the development of innovative solutions to fight COVID-19 infection. At the moment, the computer platform contains more than 800 chest X-ray examinations of COVID-19-positive patients and the clinical information of the patients, themselves.
The results of the research project are being published in the scientific journal Medical Image Analysis “AIforCOVID: predicting the clinical outcomes in patients with COVID-19 applying AI to chest-X-rays. An Italian multicenter study“. (https://arxiv.org/abs/2012.06531)
Who has joined this initiative?
In addition to the CDI, excellent hospitals and both public and private research institutes are participating:
- IRCCS Ca’ Granda Ospedale Maggiore Policlinico Foundation of Milan
- IRCCS Policlinico San Matteo Foundation of Pavia
- Azienda Ospedaliera-Universitaria Careggi of Florence
- ASST Santi Paolo e Carlo of Milan
- ASST Fatebenefratelli-Sacco of Milan
- ASST San Gerardo Hospital of Monza
Since collecting data is of little use without the option of analyzing and integrating them with each other, the contribution of the Italian Institute of Technology (IIT) in Genoa and the Campus Biomedico University in Rome has been fundamental. Together with the CDI, they’ve developed Artificial Intelligence algorithms for data analysis. An important contribution was also made by Bracco Imaging.
What are the advantages of AIforCOVID Imaging Archive and the application of Artificial Intelligence?
AIforCOVID Imaging Archive is a platform that will allow the sharing of data between research groups, in order to exponentially increase data and analysis possibilities to accelerate the development of innovative solutions against COVID-19.
In addition, algorithms based on Artificial Intelligence may be able to detect elements invisible to the human eye but useful in making a more accurate diagnosis for any worsening that may require hospitalization in intensive care (or other treatments).
The approach is therefore that of “Big Data” (click here for more details) that can be analyzed only by sophisticated algorithms able to relate the patient’s clinical data, previous diseases, radiological data and then correlate everything with the clinical development and the course of the disease.
The AIforCOVID Imaging Archive project will develop its potential only in the case of a great collaboration with other hospitals and research institutes, in order to increase the database of information and allow the algorithms to refine more and more their predictive capabilities. It will therefore be essential that other national and international clinical research institutes participate and share their data.
This post is also available in: Italiano