AI biomarker identifies aortic stenosis development, progression
An artificial intelligence-based video biomarker can accurately identify those who might develop or have rapidly worsening aortic stenosis.
Source: Getty Images
By - A multi-institutional team of researchers has identified an artificial intelligence (AI)-based video biomarker capable of helping clinicians more accurately understand which patients are likely to develop or have worsening aortic stenosis, according to a new study published in JAMA Cardiology.
Aortic stenosis – the narrowing of the aortic valve opening – is a serious disease that causes the heart to work harder to push blood throughout the body. The disease is progressive, with continued narrowing of the valve creating increased pressure in the heart and decreased oxygen supply, which can lead to potentially life-threatening complications.
Newer interventions like transcatheter aortic valve replacement can have a positive impact on disease prognosis, but the researchers emphasized that determining how aortic stenosis will progress remains a challenge. Currently, prognostic biomarkers for aortic stenosis are being researched, but these are limited in their ability to inform personalized disease screening and follow-up.
“So far, we have not had a way to know who develops aortic stenosis or who gets worse. There are no accepted clinical biomarkers for the progression of aortic stenosis, and most research to date has been focused on fixing valves once they are diseased,” said first author Evangelos K. Oikonomou, MD, DPhil, clinical fellow of cardiovascular medicine at Yale School of Medicine, in a news release. “This is foundational research that we believe will facilitate further study of new treatments to address the progression of aortic stenosis and, eventually, help prevent bad outcomes.”
The research team sought to evaluate whether there is an association between an AI video-based algorithm for aortic stenosis (AS) and the disease’s progression. In a previous study, the researchers developed a deep learning approach to identify features suggestive of aortic stenosis in cardiac ultrasound videos.
READ MORE: Machine Learning Tool Predicts Heart Failure Treatment Response
They found that a video-based biomarker, the Digital [aortic stenosis] Severity index (DASSi), was able to accurately identify the echocardiographic signature of severe aortic stenosis.
The current study looked at whether DASSi could be used to stratify the risk of aortic stenosis development and progression in patients without the condition or with mild or moderate forms of the condition at baseline.
The analysis pulled data from two cohorts of patients without severe aortic stenosis undergoing echocardiography in the Yale New Haven Health System from 2015 to 2021 and Cedars-Sinai Medical Center from 2018 to 2019.
The research team also built a novel computational pipeline to translate DASSi into cardiac magnetic resonance (CMR) imaging to assess how the biomarker would perform when applied to different imaging modalities.
Analysis of these data revealed that higher baseline DASSI was associated with faster progression in peak aortic valve velocity, and DASSi scores of 0.2 or greater were associated with a four- to five-fold higher aortic valve replacement (AVR) risk.
READ MORE: Deep Learning Tool May Help Detect Pediatric Rheumatic Heart Disease
Video features identified using DASSi are not visible to clinicians, making the tool’s ability to perform well across imaging modalities particularly valuable, the researchers emphasized.
“One of the most exciting aspects of this research is that DASSi can convert cardiac MRIs, point of care ECGs, and cardiac ultrasound all using the same model,” explained senior author Rohan Khera, MD, MS, assistant professor of cardiovascular medicine and the director of the Cardiovascular Data Science (CarDS) Lab at Yale. “This is crucial because it shows that DASSi flags a distinct myocardial and valvular phenotypic signature and is not restricted to modality-specific features or limited to some selected populations.”
The research team further indicated that DASSi could significantly improve risk stratification efforts – such as opportunistic screening – for aortic stenosis development and progression.
“I believe that every time a clinician gets a view of the heart, it’s an opportunity to screen patients and diagnose structural heart disorders,” said Khera. “This research shows that it’s possible to diagnose aortic stenosis and prognosticate risk of aortic stenosis using cardiac ultrasounds and cardiac MRI. That’s potentially practice-changing.”
Researchers from the CarDS Lab plan to validate DASSi’s diagnostic capabilities and test whether the tool can flag patients with mild or moderate aortic stenosis who may progress to a severe form of the disease.
READ MORE: AI-Driven ECG Analysis Accurately Predicts Right-Side Heart Issues
Other research is also examining the potential of advanced analytics and biomarkers to track disease progression.
Last week, a research team from Duke University School of Medicine and Pennsylvania State University showed that a personalized tool known as the Alzheimer’s Disease Biomarker Cascade (ADBC) can accurately forecast Alzheimer's disease progression.
The model uses individual biomarker data from brain scans, cerebrospinal fluid and memory tests to predict disease progression long term, compared to existing models, which use a theoretical framework to model Alzheimer’s at the molecular level over short periods.
The ADBC tool successfully identified 14 personalized parameters reflective of meaningful physiological characteristics to inform progression predictions.
