Related Experiment Video
Updated: May 16, 2025
Author Spotlight: Unraveling the Pathogenesis of Age-Related Macular Degeneration and Discovering Potential Therapies
Published on: July 28, 2023
Prediction of heart failure risk factors from retinal optical imaging via explainable machine learning
Sona M Al Younis1, Samit Kumar Ghosh1, Hina Raja2
1Department of Biomedical Engineering and Biotechnology, Healthcare Engineering Innovation Group (HEIG), Khalifa University, Abu Dhabi, United Arab Emirates.
Insights
Retinal optical coherence tomography (OCT) features show promise as non-invasive biomarkers for heart failure (HF) detection and classification. Analyzing retinal thickness may aid in early diagnosis and personalized healthcare solutions.
Area of Science:
- Ophthalmology
- Cardiology
- Biomedical Engineering
Background:
- Heart failure (HF) affects over 64 million globally, leading to high mortality and healthcare costs.
- Current diagnostic methods for HF can be invasive and costly.
- Non-invasive biomarkers are needed for early detection and classification of HF subtypes.
Purpose of the Study:
- To investigate retinal optical coherence tomography (OCT) features as non-invasive biomarkers for classifying heart failure subtypes.
- To explore the relationship between ocular indicators and heart failure using machine learning (ML).
- To assess the potential of retinal OCT for early HF detection and personalized healthcare.
Main Methods:
- Retinal measurements from optical coherence tomography (OCT) of the left eye, right eye, and both eyes were analyzed.
- Machine learning (ML) models, including logistic regression, CatBoost, and XGBoost, were employed for classification tasks.
- Nine classification experiments compared normal individuals against heart failure patients (left ventricular heart failure, congestive heart failure, unspecified heart failure).
Main Results:
- Significant reductions in retinal thickness metrics, including ISOS-RPE and macular thickness, were observed in heart failure patients.
- Machine learning models achieved robust performance, with high accuracy and AUC scores, particularly for classifying congestive heart failure and unspecified heart failure.
- Key retinal parameters like ISOS-RPE and INL-ELM thickness were identified as crucial indicators for heart failure detection.
Conclusions:
- Retinal OCT features demonstrate significant potential as non-invasive biomarkers for early detection and classification of heart failure.
- Utilizing retinal OCT data, especially from both eyes, can enhance diagnostic capabilities.
- These findings may facilitate the development of wearable diagnostic systems for improved heart failure management.
Abstract:
Over 64 million people worldwide are affected by heart failure (HF), a condition that significantly raises mortality and medical expenses. In this study, we explore the potential of retinal optical coherence tomography (OCT) features as non-invasive biomarkers for the classification of heart failure subtypes: left ventricular heart failure (LVHF), congestive heart failure (CHF), and unspecified heart failure (UHF). By analyzing retinal measurements from the left eye, right eye, and both eyes, we aim to investigate the relationship between ocular indicators and heart failure using machine learning (ML) techniques. We conducted nine classification experiments to compare normal individuals against LVHF, CHF, and UHF patients, using retinal OCT features from each eye separately and in combination. Our analysis revealed that retinal thickness metrics, particularly ISOS-RPE and macular thickness in various regions, were significantly reduced in heart failure patients. Logistic regression, CatBoost, and XGBoost models demonstrated robust performance, with notable accuracy and area under the curve (AUC) scores, especially in classifying CHF and UHF. Feature importance analysis highlighted key retinal parameters, such as inner segment-outer segment to retinal pigment epithelium (ISOS-RPE) and inner nuclear layer to the external limiting membrane (INL-ELM) thickness, as crucial indicators for heart failure detection. The integration of explainable artificial intelligence further enhanced model interpretability, shedding light on the biological mechanisms linking retinal changes to heart failure pathology. Our findings suggest that retinal OCT features, particularly when derived from both eyes, have significant potential as non-invasive tools for early detection and classification of heart failure. These insights may aid in developing wearable, portable diagnostic systems, providing scalable solutions for personalized healthcare, and improving clinical outcomes for heart failure patients.
