Aug-08-2023: In a groundbreaking development, a team of researchers led by Veer Sangha, Arash A. Nargesi, and Rohan Khera, has successfully created and externally validated a state-of-the-art deep learning model capable of detecting Left Ventricular Systolic Dysfunction (LVSD) from Electrocardiographic (ECG) images. This innovative approach offers a highly efficient and accessible screening strategy for LVSD, particularly in low-resource healthcare settings.
The research, conducted by a collaborative team of experts, involved the analysis of an astounding 385,601 ECGs along with paired echocardiograms for model development. The results were nothing short of remarkable, as the model demonstrated outstanding discrimination power across various ECG image formats and calibrations during internal validation. The area under the receiving operation characteristics (AUROC) achieved an impressive 0.91, and the area under the precision-recall curve (AUPRC) was 0.55.
Furthermore, the model's performance extended beyond internal validation, as it showcased remarkable accuracy and efficiency in identifying LVSD in external sets of ECG images sourced from prestigious institutions worldwide.
One of the most striking aspects of the study was the ability of the deep learning model to identify class-discriminative patterns localized to the anterior and anteroseptal leads (V2 to V3). Remarkably, these patterns corresponded to the left ventricle, regardless of the ECG layout, providing an unprecedented level of precision in LVSD detection.
The clinical implications of this breakthrough are immense. Researchers found that an ECG suggestive of LV systolic dysfunction indicated a staggering >27-fold higher odds of LV systolic dysfunction when confirmed through transthoracic echocardiogram. This highlights the model's potential to significantly improve the accuracy of LVSD diagnosis, leading to earlier intervention and improved patient outcomes.
Moreover, the model's impact reaches beyond the current state of LVSD. Individuals with an LV ejection fraction ≥40% at the time of initial assessment and a positive ECG screen were associated with a 3.9-fold increased risk of developing incident LV systolic dysfunction in the future. This predictive capability promises to revolutionize the way healthcare professionals monitor patients at risk and proactively intervene to prevent further complications.
The potential of this breakthrough cannot be overstated. The ability to leverage artificial intelligence to detect LVSD from routine ECGs has the potential to save countless lives and improve the quality of healthcare on a global scale.
*This press is published by VOH team.*
Hypertension is often the “silent killer” of the heart, slowly but steadily
Surge in Cardiovascular Diseases Linked to Diabetes, Hypertension, and High Chol
Apollo Hospitals Celebrates Its Leadership as the Largest Cardiac Center of Exce
Mindray Innovation Empowers Asian Games Host City Hangzhou and Beyond
Apollo Telehealth: Revolutionizing Healthcare Access Through Innovation
Dr Arvind Sharma appointed as New Executive Committee (EC) member Indian Academy
SRV Hospitals Expands its Footprint with the Launch of its Fifth Hospital in Nas
India's leading Gastroenterologist Dr Amit Maydeo joins Sir HN Reliance Foundati
Sir H. N. Reliance Foundation Hospital Takes Bold Strides to Combat Sepsis; laun
IHH Healthcare increases majority stake in India’s Gleneagles Global Hospitals
CK Birla Hospital Signs MOU with Intuitive India for launching Fourth Gen da Vin
Amrita Hospital, Faridabad, Celebrates First Anniversary of its Founding
Compassion knows no borders: At MSF's Health and Humanity Summit, a new communit
Rare Diseases Policy Summit: A DAY OF RARE INSIGHTS
ZEISS SMILE: Pioneering Vision Correction Technology Experiencing Rapid AdoptionView All