New acquisition and reconstruction solutions for ultrasound imaging developed by a Vanderbilt team aim to fundamentally improve the ability to obtain high quality, clinically relevant images, especially in cases of heart disease.
With a new, $2.3 million National Institutes of Health grant, the researchers will evaluate their advances on a specific subset of cardiac imaging used in diagnosing and monitoring patients with atrial fibrillation. Brett Byram, associate professor of biomedical engineering, and Susan Eagle, M.D., a cardiothoracic anesthesiologist at VUMC, are co-principal investigators. The team also includes Matthew Berger, assistant professor of computer science, and Chris Ellis. M.D., associate professor of medicine in the Vanderbilt Heart and Vascular Institute.
“Low-quality ultrasound is particularly problematic in echocardiography because of its role as a first-line tool for diagnosing and monitoring cardiac disease,” Byram said. “Ultrasound is the most utilized imaging modality, and echocardiography itself accounts for nearly a fifth of all imaging exams,” he said. “Unfortunately, for many diagnostic tasks, ultrasound exams are suboptimal or fail altogether.”
The project is “Ultrasound clutter and noise improvements applied to echocardiographic left atrial appendage visualization.” Imaging of the left atrial appendage, a small, ear-shaped sac in the muscle wall of the top left chamber of the heart, is growing in importance as the cases of atrial fibrillation increase. The sac is often the site of clotting that results from erratic heart rhythm.
“The methods we’re developing are being tested and evaluated in the context of left atrial appendage imaging, but they should apply to all applications of ultrasound imaging,” said Byram, who is an affiliate of the Vanderbilt Institute for Surgery and Engineering and the Vanderbilt University Institute of Imaging Science.
The team created a way transmit longer pulses that can be later compressed to boost signal-to-noise ratio without losing resolution. The primary advances are a new method for phase-based coded excitation and a new deep neural network training architecture using unlabeled in vivo data, which leads to better ultrasound image reconstruction.
Better reconstruction should support better imaging of the appendage through the chest wall without using transesophageal imaging, Byram said. A transesophageal echocardiogram involves inserting a transducer down the esophagus. It provides a clearer image of the heart but is a more invasive, more costly procedure and is not suitable for some patients.
Transthoracic imaging, which is far less expensive, involves placing the ultrasound transducer on the chest exterior.
The grant, from the NIH National Heart, Lung, and Blood Institute, will allow the team to evaluate its new methods in the context of echocardiography of the left atrial appendage. In patients with atrial fibrillation, clotting may appear in as many as 12-26% of left atria with most appearing in the appendage.
Cardiac specialists do not want to “cardiovert,” or return the heart to normal sinus rhythm, in patients with clots because the clots can dislodge and cause problems in the downstream circulatory system. Additionally, some cases of chronic atrial fibrillation are treated by implanting a device that closes the appendage, and those patients require annual transesophageal echocardiogram follow-ups. In both cases, the standard-of-care has been transesophageal echocardiography because transthoracic image quality has been considered too unreliable.
“We’ll be using transesophageal as our gold standard,” Byram said. “We’d like to improve transthoracic imaging to the point where it can be used to replace transesophageal in as many patients as possible.”