Diagnostic Imaging & Artificial Intelligence – A response to the COVID-19 Pandemic

With the prolific advancements in social media and technology, our world seems to get smaller and smaller every day. No other event in history can illustrate that point more than the global COVID-19 pandemic. As healthcare professionals we have witnessed first-hand the devastating losses caused by this pandemic but there have also been positive aspects born out of necessity that bear mentioning. A our fellow practitioners from all corners of the world have come together to share information and help each other through notably one of the most difficult times in our professions. The international medical community has band together to share findings, indications, data, what worked and what didn’t. Radiology is at the forefront of leading the international charge. Imaging modality studies such as X-ray, CT and Ultrasound combined with artificial intelligence have developed some very exciting possibilities in the fight against COVID-19.

  • Canada – A researcher, Joseph Paul Cohen, at the University of Montreal is stockpiling radiology scans to train an artificial intelligence model to recognize warning signs of severe illness. In recent weeks with the resurgence of COVID-19 in the U.S. hundreds of lung scans from clinical reports published around the world. A little more than 6 months after the pandemic emerged, Cohen along with a number of researchers and companies are already testing the ability of AI systems to aid diagnosis of COVID-19.
  • United States – The ACR is in the process of creating a repository of COVID-19 images from around the world that will post use cases on its website that may help in managing these patients
  • Italy – Aunt Minnie reported in July 2020 about a prospective study done in Italy utilizing point of care ultrasound (POCUS) on the lungs to predict which COVID-19 patients are at a greater risk of death. Physicians in Rome scanned 41 adult patients that were positive for the SARS-CoV-2 test with a standard lung scan, rating the scan with a lung ultrasound score (LUS) from 0-3 based on the findings seen during the scan.  A higher rating of 3 indicates a dense area seen. 90% of these patients had at least one area (typically in the lateral lungs) to rate the LUS as a 3.  100% of the fatal cases in this patient group had pathological findings.This indicates that ultrasound could prove to be an indicator in the severity of the disease and mortality.
  • India – Two separate studies involving artificial intelligence and imaging studies have produced stimulating possibilities in the fight against this disease.  The first is utilizing an algorithm that looks at several different areas of the lungs in a chest radiograph.  This study in its initial phase was very successful in the rapid and early identification of COVID versus other diseases of the lungs.  A second abstract indicates the success in utilizing artificial intelligence and radiology studies in CT and MRI to identify early infections, improve treatment, and health monitoring of COVID-19 patients.
  • United States – One abstract discussed how researchers in New York developed a unique AI algorithm that can rapidly diagnose COVID-19 based on the patient’s CT of the Chest in combination with symptoms, exposure history, and labs.
  • United States – At University of California-San Diego several AI teams have formed and one team is working with Amazon Web Services to train an AI to spot early signs of pneumonia, a common cause of death in COVID-19 patients.

As the complaints of inadequate, inaccurate swab testing and the amount of time it takes in some accounts to get the results back have been a common concern for discussion and complaints regarding this disease.  I find it not only interesting, but impressive that the international radiology community has not only identified uses of diagnostic imaging but within just a few months of the outbreak has published this information for others to also review.  These were just a few examples of the information already being shared for global benefits.

This is just another reminder that diagnostic imaging studies, technologists, and radiologists are making a difference everyday in the fight against COVID-19 and everyone is doing their part to improve the lives of our patients.


One thought on “Diagnostic Imaging & Artificial Intelligence – A response to the COVID-19 Pandemic

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.