LBS 03: Harnessing AI for better heart health
Improving guideline-directed medical therapy, comparing wearable versus implanted sensors, boosting echocardiogram performance and automating echocardiographic interpretation.
Saturday’s Late-Breaking Science session “Smart Cardiology: Harnessing AI and Innovation for Better Heart Health” found that:
- Pharmacy management can improve guideline-directed medical therapy for heart failure.
- Wearable sensor matches implanted sensor for pulmonary capillary wedge pressure in heart failure.
- Artificial intelligence can boost echocardiogram performance with less sonographer fatigue.
- PanEcho can automate echocardiographic interpretation.
Initial results from a randomized trial found that empowering pharmacists to titrate guideline-directed medical therapy (GDMT) for heart failure can significantly increase pharmacist engagement in heart failure care. Pharmacists embedded in primary care practices received an audit and feedback intervention in addition to heart failure education compared with usual care pharmacists that received heart failure education alone. Pharmacists in the audit and feedback arm completed more monthly heart failure encounters and more encounters with heart failure medication adjustment compared with usual care pharmacists. Heart failure patients managed by pharmacists that received audit and feedback were more likely to receive new mineralocorticoid receptor antagonist prescriptions, 11.6% versus 9.2% for usual care (p<0.01), which have historically been the most under-prescribed components of heart failure GDMT.
Alexander Sandhu, MD
“We are fortunate to have a variety of incredibly effective medication therapies that dramatically reduce morbidity and mortality of heart failure,” said Alexander Sandhu, MD, assistant professor of cardiovascular medicine at Stanford University School of Medicine. “Unfortunately, these therapies remain under-utilized. There is an urgent need to find and implement care strategies that improve our utilization of guideline-directed medical therapy for this incredibly high-risk group. “Audit and feedback” is an effective approach to increasing guideline-directed heart failure medication management by pharmacists.”
The PHARM-HF A&F Study randomized 120 primary care pharmacists in Veterans Administration facilities across Northern and Central California, Nevada and the Pacific Islands to usual care, audit and feedback, or A&F+. VA pharmacists have independent prescribing authority, Sandhu said, and are actively involved in GDMT management of hypertension, diabetes and other chronic conditions. However, pharmacists often lack the experience or comfort with managing heart failure medications.
All pharmacists in the study had access to heart failure-specific education, including a monthly webinar, to learn more about heart failure management. Pharmacists in the audit and feedback arm received emailed audit and feedback information on their heart failure medication activities compared to peers in their local VA facility and region. Pharmacists in the A&F+ arm also received lists of patients eligible for heart failure medication titration.
The primary outcome was the difference in rates of heart failure medication adjustment encounters between usual care and the combined A&F/A&F+ arms. Pharmacists’ heart failure management activities were followed for six months. An additional six months of follow-up will be reported at a later date.
“We hope that increased audit and feedback and heart failure education will empower pharmacists to confidently improve heart failure medication therapy. We hope we will see the effect of this intervention grow over time as pharmacist experience increases,” Sandhu said. “We would like to improve and adapt these practices and then scale them to other VA regions to further improve heart failure care.”
External sensor results match implanted sensor for intracardiac hemodynamics in heart failure
A novel wearable sensor that sits on the chest of patients with heart failure shows similar pulmonary capillary wedge pressure (PCWP) and other data as implanted sensors. The noninvasive CardioTag uses accelerometers to measure chest movement with each heartbeat and a novel algorithm to translate chest movement into intracardiac pressures. The SEISMIC-HF I trial showed a mean error for PCWP of 1.04 mmHg compared to gold standard right heart catheterization (RHC) measurements.
Liviu Klein, MD, MS
“A sensor on the chest is able to capture the heart’s movement in tri-dimensional space,” said Liviu Klein, MD, MS, professor of cardiology and director of the Advanced Heart Failure Comprehensive Care Center at the University of California, San Francisco. “These movements are related to heart contractions, which are related to pressures in the heart. We can get results that are very similar to an implantable device such as a CardioMEMS or Cordella PAP sensor using a wearable sensor.”
Hemodynamic-guided management of heart failure using an implantable sensor to measure pulmonary artery pressures can improve the quality of life and decrease the risk of heart failure hospitalization, Klein added. But clinical adoption has been limited due to the invasive procedure and reimbursement challenges.
The prospective observational SEISMIC-HF I study followed 943 heart failure patients at 15 U.S. centers scheduled for routine RHC. The mean age was 63, 58% were male, 55% White and 27% African American. Most, 88% had a heart failure diagnosis, 39% with LVEF ≤ 40%, and 90% had NYHA class II-IV symptoms.
In addition to the RHC, participants used a CardioTag to collect electrocardiography, seismocardiography and photoplethysmography data. The CardioTag device and algorithm received a Food and Drug Administration Breakthrough Device Designation in 2022.
The mean RHC measured pulmonary artery pressures were 42.6 mmHg systolic and 18.4 mmHg diastolic, while the mean RHC measured PCWP was 15.9 mmHg. Klein reported the validation set results showed a mean error of just over 1 mmHg for PCWP compared to gold standard RHC.
“In heart failure, we know that people feel poorly and end up in the hospital when pressures in the heart start to increase,” he said. “These pressures don’t just increase two or three days prior to hospitalization, it takes four, five, six weeks. If you have an implantable sensor, we can track pressures and adjust medications, but maybe 1% of the heart failure population has a sensor implanted. This noninvasive device can provide similar information that can prevent patients from ending up in the hospital, prevent them from having symptoms and help them lead a more normal daily life at home.”
Artificial intelligence can improve echocardiographic workflow
A single-center randomized crossover study found that a novel artificial intelligence (AI)-based analysis tool can streamline the daily workflow in echocardiology with improved measurements and more patients examined compared to conventional manual echocardiography for cardiovascular risk assessment. AI assistance reduced the time per exam, 13.0 minutes versus 14.3 minutes for manual exams (p<0.001) and increased the number of daily exams from 14.1 to 16.7 (p=0.003) with less sonographer fatigue (p=0.039), 3.4-fold more echocardiographic parameters analyzed per exam (85 versus 25, p<0.001) and improved cardiographic image quality (p<0.001).
“Over 90% of the AI’s initial values were clinically acceptable and used in clinical practice,” said Nobuyuki Kagiyama, MD, PhD, associate professor of cardiology at Juntendo University School of Medicine in Tokyo, Japan. “AI can enhance efficiency in the echo lab, easing a boring, repetitive task like screening echocardiograms, so sonographers and cardiologists can spend more time on the detailed evaluation of more severe patients who really need more intensive care and attention.”
Improving echocardiography workflow is a particular interest in Japan, which has about 1/3 the U.S. population but performs about 1.3 times more echocardiograms, commonly for routine cardiovascular risk screening. AI-ECHO randomized four experienced sonographers performing screening echocardiography over 38 days on a daily basis to use AI for automatic echocardiography analysis (19 AI days) or conventional procedures (19 non-AI days). Both AI and non-AI echocardiograms were reviewed by expert cardiologists, who finalized all reports for clinical use.
The primary endpoint was examination efficiency, defined as the time per examination and the number of exams performed per day. Secondary endpoints included the number of parameters analyzed and image quality.
Nobuyuki Kagiyama, MD, PhD
AI days allowed sonographers to focus on image acquisition and quality, Kagiyama said, resulting in an overall improvement in image quality compared to non-AI days. Because AI was handling image analysis, sonographers could, and did, concentrate more on acquiring higher quality images knowing that they would not have to spend time later evaluating imaging themselves.
“This software is already approved by the FDA and the Pharmaceuticals and Medical Devices Agency in Japan for some uses, but AI-ECHO pushed it beyond what is approved today,” Kagiyama said. “This real-world randomized trial demonstrates how AI-based automatic analysis can significantly improve the efficiency of screening echocardiography by reducing exam time while maintaining image quality and reducing sonographer fatigue.”
AI interpretation can accurately interpret echocardiogram findings across multiple metrics
Transthoracic echocardiography (TTE) is a key tool for cardiovascular evaluation, but manual reporting can be slow, and interpretation is subject to intra-reviewer variability. A novel AI tool, PanEcho, is the first view-agnostic, multitask AI model that automates TTE interpretation across views and acquisitions for all key echocardiographic metrics and findings. An initial validation study showed a median area under the receiver operating characteristic curve (AUC) of 0.91 across 18 classifications. Key findings include an AUC of 0.99 to detect severe aortic stenosis, 0.98 for moderate-severe left ventricular (LV) systolic dysfunction and 0.95 for moderate-severe LV dilation.
“To our knowledge, this is the first AI model to provide comprehensive echocardiogram interpretation from multiview echocardiography,” said Gregory Holste, MSE, graduate student at the Yale School of Medicine Cardiovascular Data Science (CarDS) Lab. “Current AI applications in echocardiography have been limited to single views and single pathologies for outcomes. And intrepetation is nearly real time."
Gregory Holste, MSE
PanEcho was developed using 1.23 million echocardiographic videos from 33,927 TTE studies performed at a New England health system between January 2016 and June 2022. The model can perform 39 TTE reporting tasks spanning the full spectrum of myocardial and valvular structure and function from parasternal, apical and subcostal views, including B-mode and color Doppler videos. The model was evaluated on a distinct New England health system cohort and two cohorts in California. Researchers assessed off-the-shelf diagnostic performance and PanEcho’s ability to function as a foundational model that can be fine-tuned for specific domains.
Rohan Khera, MD, MS
The model estimated continuous metrics with a median normalized mean absolute error (MAE) of 0.13 across 21 routine echocardiographic tasks. LV ejection fraction (EF) can be estimated with 4.4% MAE and LV internal diameter with 3.8 mm MAE. PanEcho can identify which views are most informative for each task. The model has transferred LVEF estimation to novel pediatric populations with superior performance compared to existing approaches, 3.9% MAE versus 4.5% MAE for the next-best approach.
“We see strong predictive performance even in very simplified acquisition of just five videos from key views,” said principal investigator Rohan Khera, MD, MS, CarDS Director. “Such applications to simpler acquisitions could broaden the efficient, expert-level interpretation of PanEcho even to point-of-care ultrasound, especially suited to low-resource settings. The next step is prospective validation in a real-world clinical workflow.”
