AI Detects Hidden Cardiac Damage in Queens Patient, Prompting Early Intervention

In February 2025, Louie Quiros, a 45-year-old caregiver and security guard, arrived at a Queens emergency department coughing up blood, experiencing worsening shortness of breath, and reporting a rapid heartbeat. Standard imaging (chest X-ray) showed no abnormality, and an electrocardiogram (ECG) was flagged as abnormal but did not point to a clear diagnosis. After learning that Quiros had recently been exposed to wildfire smoke, clinicians initially prescribed asthma medication and an inhaler. As part of a clinical trial, the hospital’s AI system, EchoNext, automatically re-examined his ECG and identified patterns suggestive of serious cardiac injury that human readers had missed. The patient was recalled a week later for an echocardiogram, which confirmed a critically low ejection fraction—only 10 percent of blood pumped per beat—and a leaking mitral valve. The early AI alert led to rapid cardiology referral, a step clinicians described as potentially life-saving.

New York Times

EchoNext is an artificial-intelligence program created within the NewYork-Presbyterian health system and Columbia University Irving Medical Center. The tool was designed to scan ECGs for subtle signatures of heart damage that are not readily apparent to clinicians. Researchers deployed the algorithm across the system’s ECG database, analyzing nearly 500,000 recordings each year. The AI processes each ECG in under ten minutes, delivering its assessment to the ordering clinician shortly after the test is performed.

New York Times

Dr. Pierre Elias, medical director of AI at NewYork-Presbyterian and cardiologist at Columbia University Irving Medical Center, oversaw EchoNext’s development. He has founded Pathway Labs to commercialize the technology beyond the academic setting. Dr. Elias emphasizes that the algorithm “reads an ECG in less than ten minutes and evaluates patterns that a human eye might overlook,” positioning the system as a decision-support aid rather than a replacement for physician judgment.

New York Times

• Annual ECG volume: ~500,000 recordings processed by the AI.

New York Times

• Turnaround time: <10 minutes from ECG acquisition to AI output.

New York Times

• Clinical finding in Quiros: Ejection fraction ? 10 % and mitral regurgitation detected after AI-triggered recall.

New York Times

NewYork-Presbyterian’s research team reported that the AI’s flagging of Quiros’ ECG “prompted a follow-up echocardiogram that revealed severe systolic dysfunction, a condition that would likely have progressed unnoticed.” Dr. Elias added that the ongoing clinical trial aims to quantify how often EchoNext uncovers hidden cardiac pathology compared with standard interpretation. The hospital has not yet released broader outcome data from the trial.

New York Times

The Quiros case illustrates how AI-augmented ECG review can identify life-threatening cardiac impairment even when conventional tests appear inconclusive. Early detection of low ejection fraction and valve leakage enables timely referral for advanced therapies, potentially reducing morbidity and mortality among patients whose symptoms are atypical or masked by other conditions (e.g., smoke inhalation).

New York Times

Pathway Labs is preparing to market EchoNext to additional health systems pending regulatory clearance. The clinical trial continues to enroll patients across NewYork-Presbyterian facilities, with investigators planning to publish comparative performance metrics later in 2026. If the trial confirms the algorithm’s diagnostic yield, broader adoption could reshape emergency-room cardiac triage protocols nationwide.

New York Times