Northwestern University Has Developed An AI System That Helps Detect Covid-19 On Chest X-Rays

Earlier last week, Northwestern University researchers announced that they successfully created a new Artificial Intelligence (AI) radiology tool that can detect Covid-19 in chest x-rays.

The study has since been published in the journal Radiology, and indicates that the system “classified 2,214 test images with an accuracy of 83%.”

Dr. Aggelos Katsaggelos, a senior author of the study, states in the press report that “We are not aiming to replace actual testing […] 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.” Dr. Ramsey Wehbe, another main author of the study, explained that “It could take hours or days to receive results from a COVID-19 test […] A.I. 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.”

As Katsaggelos so aptly describes, the ability to conduct an initial screening to see if patients need to be isolated could itself be a potentially massive value addition to emergency department physicians. During the height of the pandemic, and still in many places, personal protective equipment (PPE) was one of the first supplies to run low, meaning that healthcare professionals were routinely seeing coronavirus positive patients without protection for themselves, potentially exacerbating the spread of the virus. In fact, this caused many healthcare workers to often reuse and stretch out limited supplies of PPE for patient care. Per the Centers for Disease Control and Prevention (CDC), so far, nearly 238,000 healthcare professionals have contracted Covid-19, with over 841 having passed away due to the virus.

The discussion in the journal article also provides an important consideration of this technology: “Prior clinical studies showed COVID-19 pneumonia produces characteristic features on chest imaging, but up to 56% of symptomatic patients can demonstrate normal chest imaging, especially early in their disease course. Imaging is therefore inappropriate to “rule out” disease. Also, many of the findings seen in COVID-19 imaging are non-specific with overlap, particularly with other viral pneumonias. Chest imaging therefore should not be used as a diagnostic tool for COVID-19, but could play an important role in earlier identification of patients likely to have the disease to aid in triage and infection control.”

The press report does also warn that “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 these cases, this AI radiology tool will likely not be very helpful.

Nonetheless, as the authors of the study so aptly conclude: “We feel that this algorithm has the potential to benefit healthcare systems in mitigating unnecessary exposure to the virus by serving as an automated tool to rapidly flag patients with suspicious chest imaging for isolation and further testing.” Indeed, if this screening tool continues to be further tested, and can be proven to be efficacious, safe, viable, and somewhat scalable, it may potentially be able to help alleviate some of the burden that healthcare workers face.

The content of this article is not implied to be and should not be relied on or substituted for professional medical advice, diagnosis, or treatment by any means, and is not written or intended as such. This content is for information and news purposes only. Consult with a trained medical professional for medical advice.