Artificial Intelligence Tool Identifies COVID-19 in CT Scans

An artificial intelligence tool can quickly detect COVID-19 based on CT scans of the chest and patients’ clinical data, according to a study published in Nature Medicine.

For more coronavirus updates, visit our resource page, updated twice daily by Xtelligent Healthcare Media.

The standard way of testing for COVID-19 can take up to two days to complete, researchers noted. Chest CT scans are a useful tool for evaluating and diagnosing symptomatic patients suspected to have the virus, but these scans alone have limited predictive value. This highlights the need to incorporate clinical data into the diagnosis process – an ideal use case for artificial intelligence.

"AI has huge potential for analyzing large amounts of data quickly, an attribute that can have a big impact in a situation such as a pandemic,” said Zahi Fayad, PhD, Director of the BioMedical Engineering and Imaging Institute (BMEII) at the Icahn School of Medicine at Mount Sinai.  

“At Mount Sinai, we recognized this early and were able to mobilize the expertise of our faculty and our international collaborations to work on implementing a novel AI model using CT data from coronavirus patients in Chinese medical centers.”

The team used scans of more than 900 patients that they had received from institutional collaborators at hospitals in China. The scans included 419 confirmed COVID-19-poisitve cases and 486 COVID-19-negative scans.

Researchers also leveraged patients’ clinical information, including blood test results showing any abnormalities in white blood cell counts or lymphocyte counts, as well as their age, sex, and symptoms. They focused on CT scans and blood tests because doctors in China use both of these to diagnose patients with COVID-19.

The group then integrated data from those CT scans with the clinical information to develop an AI algorithm. The tool mimics the workflow a physician uses to diagnose COVID-19 and gives a final prediction of positive or negative diagnosis. The model also produces separate probabilities of being COVID-19-positive based on CT scans, clinical data, and both combined.

Initially, researchers trained and fine-tuned the algorithm on data from 626 out of 905 patients, and then tested the model on the remaining 279 patients in the study group to judge the test’s sensitivity. The results showed that the algorithm had statistically significantly higher sensitivity, at 84 percent, compared to radiologists’ sensitivity of 75 percent when evaluating the images and clinical data.

The algorithm also improved the detection of COVID-19-positive patients who had negative CT scans: It recognized 68 percent of COVID-19-positive cases, while radiologists interpreted all these cases as negative because of the negative CT appearance.

"We were able to show that the AI model was as accurate as an experienced radiologist in diagnosing the disease, and even better in some cases where there was no clear sign of lung disease on CT," said Fayad, who is also a Professor of Diagnostic, Molecular and Interventional Radiology at the Icahn School of Medicine at Mount Sinai.

"We're now working on how to use this at home and share our findings with others--this toolkit can easily be deployed worldwide to other hospitals, either online or integrated into their own systems."

Researchers also pointed out that improved detection is particularly important to keep patients isolated if scans don’t show lung disease when patients first present symptoms. Additionally, COVID-19 symptoms are often nonspecific, so it can be difficult to diagnose.

While CT scans are not widely used for diagnosis in the US, imaging can still play a critical role.

"Imaging can help give a rapid and accurate diagnosis--lab tests can take up to two days, and there is the possibility of false negatives--meaning imaging can help isolate patients immediately if needed, and manage hospital resources effectively,” said Fayad.  

“The high sensitivity of our AI model can provide a 'second opinion' to physicians in cases where CT is either negative (in the early course of infection) or shows nonspecific findings, which can be common. It's something that should be considered on a wider scale, especially in the United States, where currently we have more spare capacity for CT scanning than in labs for genetic tests.”

In future research, the team will continue to refine the model to discover clues about how well patients will do based on subtleties in their CT data and clinical information. This approach could be used to improve COVID-19 treatment and outcomes, as well as mitigate the spread of the virus.

"This study is important because it shows that an artificial intelligence algorithm can be trained to help with early identification of COVID-19, and this can be used in the clinical setting to triage or prioritize the evaluation of sick patients early in their admission to the emergency room," said Matthew Levin, MD, Director of the Mount Sinai Health System's Clinical Data Science Team, and a member of the Mount Sinai COVID Informatics Center.

"This is an early proof concept that we can apply to our own patient data to further develop algorithms that are more specific to our region and diverse populations."