Distinguishing heart failure with reduced ejection fraction (HFrEF) from heart failure with preserved ejection fraction (HFpEF) is clinically important but challenging. This paper presents an end-to-end deep learning pipeline for automated three-class classification (healthy, HFrEF, HFpEF) from 2D apical four-chamber echocardiographic videos, combining the public RVENet dataset with institutional data from the University of Louisville. After a multi-step preprocessing pipeline (colour-space conversion, cardiac-cycle detection, motion-based cropping, normalization), a lightweight 3D SqueezeNet and a deeper ResNet50 were evaluated as spatio-temporal classifiers, achieving AUC values of 0.93-0.98 across classes, with ResNet50 reaching an F1-score of up to 0.9205 for HFrEF when using two cardiac cycles as input. This work was presented at SPIE Medical Imaging 2026: Clinical and Biomedical Imaging, Vancouver, BC, Canada, and was carried out in part within the STRATIFYHF project.
Voice characteristics are an emerging, non-invasive biomarker for heart failure. This study develops a machine learning pipeline to differentiate patients with suspected heart failure from those with a confirmed diagnosis using vocal features alone, drawing on 240 patients (50 su…
Heart failure presents with a wide range of symptoms that affect patients' quality of life. This study uses physical-examination data and blood biomarkers to predict the emergence of thirteen individual HF-related symptoms (e.g. dyspnea, orthopnea, peripheral oedema, pulmonary cr…
Morbidity and disease severity in heart failure are commonly assessed using New York Heart Association (NYHA) classes. This study uses non-invasive data — physical examination, symptoms and disease history — to classify patients into four morbidity classes approximati…
Accurate classification of echocardiographic views (e.g. apical 2- and 4-chamber, parasternal long-axis) is an important prerequisite for reliable ejection-fraction assessment and heart failure diagnosis, but manual classification is time-consuming. This study evaluates the EchoJ…
Heart failure and chronic obstructive pulmonary disease often present with overlapping clinical signs, making differential diagnosis challenging. This work applies machine learning, primarily random forest ensembles trained on heterogeneous clinical data (physical examination, bl…
Heart failure is one of the most life-threatening diseases of the modern era, with high global mortality and morbidity rates, motivating the need for long-term outcome prediction. One established tool is the MAGGIC Risk Calculator for Heart Failure, which predicts 3-5 year mortal…