By Nancy Lapid
(Reuters) -Artificial intelligence (AI) can turn a common doctor’s office test into a screening tool for detecting structural problems in patients’ hearts, researchers reported in Nature.
Their publicly available AI tool, called EchoNext, analyzes ordinary electrocardiogram (ECG) data to identify patients who should have an echocardiogram – a noninvasive ultrasound exam – to look for valve diseases, thickening of the muscle tissue, and other structural defects that can impair heart function.
“We were all taught in medical school that you can’t detect structural heart disease from an electrocardiogram,” study leader Pierre Elias of Columbia University Vagelos College of Physicians and Surgeons said in a statement.
“We think that ECG plus AI has the potential to create an entirely new screening paradigm.”
EchoNext uses the cheaper ECG to figure out who needs the more expensive ultrasound, he said.
When 13 cardiologists reviewed a total of 3,200 ECGs, they detected structural heart problems with an accuracy rate of about 64%, compared to a 77% accuracy rate for EchoNext, the researchers found.
They next used the tool to review ECGs obtained in the past from nearly 85,000 patients. Based on those ECGs, the patients’ doctors had sent 4,100 of them to get echocardiograms, which found structural problems in roughly 3,000. But EchoNext identified an additional 3,400 patients as being at high risk and needing the ultrasound exam.
Because AI was unavailable when those ECGs were obtained, many of the additional patients may have had potentially serious structural heart disease that went undiagnosed, the researchers said.
“You can’t treat the patient you don’t know about,” Elias said. “Using our technology, we may be able to turn the estimated 400 million ECGs that will be performed worldwide this year into 400 million chances to screen for structural heart disease and potentially deliver life-saving treatment at the most opportune time.”
Worldwide, structural heart disease impacts 64 million people with heart failure and 75 million with valvular disease, with costs in the U.S. alone exceeding $100 billion annually, the researchers said.
SOME BRAIN CELLS HAVE BACKUP BATTERIES
Neurons, the nerve cells that transmit information to and from the brain, are equipped with “backup batteries” that kick in to keep the brain running during periods of metabolic stress, researchers have discovered.
Traditionally, it was believed that brain cells called glial cells served as “energy warehouses” for the neurons, storing a form of sugar known as glycogen and supplying it as needed for fuel.
“But we now know that neurons themselves store glycogen and can break it down when the pressure is on,” study leader Milind Singh of the Yale School of Medicine said in a statement.
“It’s like discovering that your car is a hybrid — it’s not just reliant on gas stations, it’s been carrying an emergency battery the whole time.”
Their discovery was made during experiments with a microscopic roundworm called C. elegans and a fluorescent sensor that glows when cells break down sugar for energy.
The findings could shape new treatments for neurological conditions in which energy failure plays a role, such as stroke, neurodegeneration, and epilepsy, the researchers said in PNAS.
The team found the neuron’s glycogen-dependent energy production is especially important when their mitochondria – their primary energy factories – are impaired, such as when the oxygen supply is limited.
Under these conditions, glycogen serves as a rapid-access fuel source, helping neurons stay active when other systems might stall, the researchers said.
“That flexibility might be crucial for how the brain maintains function and responds to stress,” senior researcher Daniel Colón-Ramos, also of Yale, said in a statement.
“This research reshapes our understanding of brain energy metabolism and opens new avenues for exploring how to protect and support neuronal function in disease.”
(To receive the full newsletter in your inbox for free sign up here)
(Reporting by Nancy Lapid; additional reporting by Shawana Alleyne-Morris; editing by Aurora Ellis)
Comments