News | Artificial Intelligence | December 03, 2020

AI Detects COVID-19 on Chest X-rays with Accuracy and Speed

System could also potentially flag patients for isolation and testing who are not otherwise under investigation for COVID-19

Generated heatmaps appropriately highlighted abnormalities in the lung fields in those images accurately labeled as COVID-19 positive (A-C) in contrast to images which were accurately labeled as negative for COVID-19 (D). Intensity of colors on the heatmap correspond to features of the image that are important for prediction of COVID-19 positivity. Image courtesy of Northwestern University

Generated heatmaps appropriately highlighted abnormalities in the lung fields in those images accurately labeled as COVID-19 positive (A-C) in contrast to images which were accurately labeled as negative for COVID-19 (D). Intensity of colors on the heatmap correspond to features of the image that are important for prediction of COVID-19 positivity. Image courtesy of Northwestern University

December 3, 2020 — Northwestern University researchers have developed a new artificial intelligence (AI) platform that detects COVID-19 by analyzing X-ray images of the lungs.

Called DeepCOVID-XR, the machine-learning algorithm outperformed a team of specialized thoracic radiologists — spotting COVID-19 in X-rays about 10 times faster and 1-6% more accurately.

The researchers believe physicians could use the AI system to rapidly screen patients who are admitted into hospitals for reasons other than COVID-19. Faster, earlier detection of the highly contagious virus could potentially protect health care workers and other patients by triggering the positive patient to isolate sooner.

The study's authors also believe the algorithm could potentially flag patients for isolation and testing who are not otherwise under investigation for COVID-19.

The study was published on Nov. 24 in the journal Radiology.

"We are not aiming to replace actual testing," said Northwestern's Aggelos Katsaggelos, an AI expert and senior author of the study. "X-rays are routine, safe and inexpensive. It would take seconds for our system to screen a patient and determine if that patient needs to be isolated."

"It could take hours or days to receive results from a COVID-19 test," said Ramsey Wehbe, M.D., a cardiologist and postdoctoral fellow in AI at the Northwestern Medicine Bluhm Cardiovascular Institute. "AI doesn't confirm whether or not someone has the virus. But if we can flag a patient with this algorithm, we could speed up triage before the test results come back."

Katsaggelos is the Joseph Cummings Professor of Electrical and Computer Engineering in Northwestern's McCormick School of Engineering. He also has courtesy appointments in computer science and radiology. Wehbe is a postdoctoral fellow at Bluhm Cardiovascular Institute at Northwestern Memorial Hospital.

A trained eye

For many patients with COVID-19, chest X-rays display similar patterns. Instead of clear, healthy lungs, their lungs appear patchy and hazy.

"Many patients with COVID-19 have characteristic findings on their chest images," Wehbe said. "These include 'bilateral consolidations.' The lungs are filled with fluid and inflamed, particularly along the lower lobes and periphery."

The problem is that pneumonia, heart failure and other illnesses in the lungs can look similar on X-rays. It takes a trained eye to tell the difference between COVID-19 and something less contagious.

Katsaggelos' laboratory specializes in using AI for medical imaging. He and Wehbe had already been working together on cardiology imaging projects and wondered if they could develop a new system to help fight the pandemic.

"When the pandemic started to ramp up in Chicago, we asked each other if there was anything we could do," Wehbe said. "We were working on medical imaging projects using cardiac echo and nuclear imaging. We felt like we could pivot and apply our joint expertise to help in the fight against COVID-19."

AI vs. human

To develop, train and test the new algorithm, the researchers used 17,002 chest X-ray images -- the largest published clinical dataset of chest X-rays from the COVID-19 era used to train an AI system. Of those images, 5,445 came from COVID-19-positive patients from sites across the Northwestern Memorial Healthcare System.

The team then tested DeepCOVID-XR against five experienced cardiothoracic fellowship-trained radiologists on 300 random test images from Lake Forest Hospital. Each radiologist took approximately two-and-a-half to three-and-a-half hours to examine this set of images, whereas the AI system took about 18 minutes.

The radiologists' accuracy ranged from 76-81%. DeepCOVID-XR performed slightly better at 82% accuracy.

"These are experts who are sub-specialty trained in reading chest imaging," Wehbe said. "Whereas the majority of chest X-rays are read by general radiologists or initially interpreted by non-radiologists, such as the treating clinician. A lot of times decisions are made based off that initial interpretation."

"Radiologists are expensive and not always available," Katsaggelos said. "X-rays are inexpensive and already a common element of routine care. This could potentially save money and time — especially because timing is so critical when working with COVID-19."

Limits to diagnosis

Of course, not all COVID-19 patients show any sign of illness, including on their chest X-rays. Especially early in the virus' progression, patients likely will not yet have manifestations on their lungs.

"In those cases, the AI system will not flag the patient as positive," Wehbe said. "But neither would a radiologist. Clearly there is a limit to radiologic diagnosis of COVID-19, which is why we wouldn't use this to replace testing."

The Northwestern researchers have made the algorithm publicly available with hopes that others can continue to train it with new data. Right now, DeepCOVID-XR is still in the research phase, but could potentially be used in the clinical setting in the future.

For more information: www.northwestern.edu

Related Content

#coronavirus #COVID19 #pandemic

Getty Images

News | Radiology Imaging | January 01, 2021
The Imaging Technology News (ITN) team wishes you a Happy and Healthy New Year!
Now, researchers using magnetic resonance imaging (MRI) have found potential brain biomarkers of PTSD in people with traumatic brain injury (TBI).
News | Magnetic Resonance Imaging (MRI) | December 31, 2020
December 31, 2020 — Posttraumatic stress disorder (...
After “COVID-19,” the term that most people will remember best from 2020 is likely to be “social distancing.” While it most commonly applied to social gatherings with family and friends, it has impacted the way many receive medical care. Historically, the United States has been relatively slow to broadly adopt telemedicine, largely emphasizing in-person visits.

Getty Images

News | Teleradiology | December 31, 2020
December 31, 2020 — After “COVID-19,” the term t
Differences in brain activity between connected and disconnected states of consciousness studied with positron emission tomography (PET) imaging. Activity of the thalamus, anterior (ACC) and posterior cingulate cortices (PCC), and bilateral angular gyri (AG) show the most consistent associa-tions with the state of consciousness (A = general anesthesia, B = sleep). The same brain struc-tures, which are deactivated when the state of consciousness changes to disconnected in general anesthesia or natural sleep

Differences in brain activity between connected and disconnected states of consciousness studied with positron emission tomography (PET) imaging. Activity of the thalamus, anterior (ACC) and posterior cingulate cortices (PCC), and bilateral angular gyri (AG) show the most consistent associa-tions with the state of consciousness (A = general anesthesia, B = sleep). The same brain struc-tures, which are deactivated when the state of consciousness changes to disconnected in general anesthesia or natural sleep (cool colors in the left columns), are reactivated when regaining a con-nected state upon emergence from anesthesia (warm colors in the right columns). Graphic courtesy of University of Turku

News | PET Imaging | December 30, 2020
December 30, 2020 — What happens in the brain when our conscious awareness fades during general anesthesia and normal
Volpara Health announced two new research studies using AI-powered software to score breast density objectively and consistently to evaluate its impact in mammography and breast cancer risk assessment.
News | Breast Density | December 30, 2020
December 30, 2020 — Volpara Health announced two new research studies using...
Researchers at the University of Turku have discovered what type of neural mechanisms are the basis for emotional responses to music. Altogether 102 research subjects listened to music that evokes emotions while their brain function was scanned with functional magnetic resonance imaging (fMRI). The study was carried out in the national PET Centre.

Getty Images

News | Magnetic Resonance Imaging (MRI) | December 30, 2020
December 30, 2020 — Researchers at the University of Turku have di
This novel agent, based on the essential nutrient manganese, has strong potential as a safer alternative to currently used contrast agents
News | Contrast Media | December 28, 2020
December 28, 2020 — Adding a contrast-enhancing agent to...
All but 3 of the top 25 articles in 2020 were on COVID-19. The other 3 were on MRI contrast advances and the development of a cold-cathode X-ray tube technology. #COVID19

All but 3 of the top 25 articles in 2020 were on COVID-19. The other 3 were on MRI contrast advances and the development of a cold-cathode X-ray tube technology.

Feature | December 23, 2020 | Dave Fornell, Editor
Here are the top 25 best performing articles on the Imaging Technology News (ITN) website from 2020.
The top two videos on ITN for the year both involved medical imaging of COVID using point of care ultrasound (POCUS) and mass movement to teleradiology to enable remote working for radiologists and virtual collaboration with referring physicians. The image on the left is Butterfly's POCUS system that turns a smart phone into an ultrasound machine and the image is of COVID B-lines in the lung. The image on the right is a CT scan of COVID pneumonia. #COVID19

The top two videos on ITN for the year both involved medical imaging of COVID using point of care ultrasound (POCUS) and mass movement to teleradiology to enable remote working for radiologists and virtual collaboration with referring physicians. The image on the left is Butterfly's POCUS system that turns a smart phone into an ultrasound machine and the image is of COVID B-lines in the lung. The image on the right is a CT scan of COVID pneumonia. 

Feature | December 23, 2020 | Dave Fornell, Editor
Here are the top 25 best performing videos on the Imaging Technology News website (ITN) from the past year, based on
Company delivers on last year’s roadmap milestones and continues to advance cloud-native suite of tools to lead industry to the future of enterprise imaging
News | Enterprise Imaging | December 23, 2020
December 23, 2020 — ...