Chatbot News, Emerging Technology
How AI Is Helping with More Accurate Diagnosing During COVID-19
Artificial intelligence (AI) is getting increasingly better at mimicking and performing what humans can do—sometimes even in more efficient and affordable ways than what we are accustomed to. The potential for both AI and robotics in healthcare is vast. Just like in our everyday lives, AI and robotics are increasingly a part of our healthcare ecosystem.Let’s take a look at the most recent discoveries on how AI is helping to more accurately diagnose health ailments during the COVID-19 pandemic.
By Tauren Dyson
October 20, 2020
As COVID-19 continues to spread, science continues to reveal the negative impacts the virus has on the heart. A new artificial intelligence solution may help to quickly and accurately diagnose the cause of heart failure better than traditional methods used to detect the condition.
Those results come from a study recently published in Circulation: Arrhythmia and Electrophysiology, an American Heart Association journal.
Prior to the pandemic, about 1.2 million people visited emergency departments each year due to shortness of breath—a common symptom of COVID-19. Since the onset of the pandemic, those numbers have climbed, since breathing difficulty is the main symptom of a COVID-19 infection.
Since that time, the U.S. Food and Drug Administration has authorized the emergency use of new AI-enhanced ECGs to diagnose heart failure in people who have contracted or are suspected of having COVID-19.
Some research has shown that coronavirus can trigger death more quickly in people with heart failure and other cardiovascular conditions.
Shortness of breath is also one of the key warning signs of heart failure. But just having breathing problems doesn’t mean a patient has heart failure.
That’s why it’s important that physicians have a tool to improve their accuracy in identifying these cardiovascular events.
"Determining why someone has shortness of breath is challenging for emergency department physicians, and this AI-enabled ECG provides a rapid and effective method to screen these patients for left ventricular systolic dysfunction," Demilade Adedinsewo, MD, MPH, the study’s lead author and a chief fellow in the division of cardiovascular medicine at Mayo Clinic in Jacksonville, Florida, told the American Cancer Society.
Physicians perform ECG tests on patients suspected of having heart problems. It’s a 10-second test that records the heart’s electrical activity.
But the test alone can’t determine if a patient has suffered heart failure.
"An abnormal ECG raises concern about underlying cardiac abnormalities but is not specific for heart failure," Adedinsewo said.
To detect whether a patient has suffered heart failure, doctors also typically measure blood levels of natriuretic peptides. Heart failure patients often have high levels of peptides in their blood.
Risk factors that raise these biomarker levels include age, abnormal heart rhythms, kidney disease, obesity, severe infection, and high blood pressure.
Researchers from Mayo Clinic set out to develop an AI solution that could beat the accuracy of a traditional ECG.
They created an AI-enhanced ECG by training it to distinguish the health records of patients diagnosed with heart failure from those without the condition.
They taught the AI-enhanced ECG to analyze medical records for more than 1,600 patients who were tested with a traditional ECG and blood testing during emergency room visits.
To do this, the researchers used the AI-enhanced ECG to pinpoint in patients visiting the emergency room with shortness of breath in May 2018 and February 2019.
The researchers found that the AI solution diagnosed severe heart failure in nearly 90 percent of the patients. That’s compared to only 80 percent using the traditional ECG.
The AI-enhanced solution also identified low pumping ability in roughly 85 percent of patients.
"AI-enhanced ECGs are quicker and outperform current standard-of-care tests. Our results suggest that high-risk cardiac patients can be identified quicker in the emergency department and provides an opportunity to link them early to appropriate cardiovascular care," Adedinsewo said.
AI is also making advances in treating pulmonary problems brought on by coronavirus. Researchers at the University of Central Florida (UCF) have developed an algorithm that can diagnose coronavirus in the lungs with nearly the same accuracy as a doctor.
Researchers think this AI solution can only enhance some of the inherent flaws in the traditional method. Typically, doctors use a CT scan to pinpoint the virus in the lungs, but this can also cause misreads.
Some CT scans can’t distinguish between coronavirus and the flu.
The UCF team says they have created an algorithm to identify COVID-19 with a precision better than a CT scan. The study was published in Nature Communications.
For the study, researchers taught a computer algorithm to recognize COVID-19 in CT scans for nearly 1,300 patients from China, Japan, and Italy who were suffering from the condition.
Then the researchers turned the algorithm on more than 1,300 patients with lung diseases, such as COVID-19, cancer, and pneumonia which wasn’t brought on by COVID-19. After comparing the results to the diagnoses delivered by doctors using traditional methods, 90 percent of the COVID-19 cases identified by the AI-solution were accurate. The AI-solution also identified with accuracy positive COVID-19 cases in 84 percent of patients and negative cases 93 percent.
Ulas Bagci works as an assistant professor at UCF’s Department of Computer Science and is a co-author of the study. He’s confident this AI solution to detect COVID-19 in a patient’s lungs can be used to help diagnose, prevent, track, and research the virus.
“We demonstrated that a deep learning-based AI approach can serve as a standardized and objective tool to assist healthcare systems as well as patients. It can be used as a complementary test tool in very specific limited populations, and it can be used rapidly and at a large scale in the unfortunate event of a recurrent outbreak.”