Predicting Pulmonary Hypertension in Newborns: A Multi-view VAE Approach

2025-07-14

AI Summary

Current non-invasive diagnosis of neonatal pulmonary hypertension (PH) relies heavily on operator expertise and suffers from poor generalizability due to single-view echocardiography. Method: We propose an unsupervised representation learning framework based on a multi-view variational autoencoder (VAE), the first to model temporal dynamics across multiple standard echocardiographic views (e.g., PSAX, AP4C, AP2C) for neonatal PH prediction. The model jointly learns latent spatiotemporal features from synchronized video sequences, enabling cross-view complementary information fusion and robust representation extraction. Contribution/Results: Evaluated on a real-world neonatal cohort, our method significantly outperforms both single-view and supervised baseline models—achieving a 6.2% absolute improvement in classification accuracy—and demonstrates superior cross-center generalizability. This work establishes a novel, interpretable, operator-agnostic, and robust AI-assisted diagnostic paradigm for pediatric cardiac functional assessment.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Computer Vision: Multi-modal Vision

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Pulmonary hypertension (PH) in newborns is a critical condition characterized by elevated pressure in the pulmonary arteries, leading to right ventricular strain and heart failure. While right heart catheterization (RHC) is the diagnostic gold standard, echocardiography is preferred due to its non-invasive nature, safety, and accessibility. However, its accuracy highly depends on the operator, making PH assessment subjective. While automated detection methods have been explored, most models focus on adults and rely on single-view echocardiographic frames, limiting their performance in diagnosing PH in newborns. While multi-view echocardiography has shown promise in improving PH assessment, existing models struggle with generalizability. In this work, we employ a multi-view variational autoencoder (VAE) for PH prediction using echocardiographic videos. By leveraging the VAE framework, our model captures complex latent representations, improving feature extraction and robustness. We compare its performance against single-view and supervised learning approaches. Our results show improved generalization and classification accuracy, highlighting the effectiveness of multi-view learning for robust PH assessment in newborns.

Problem

• Predicting pulmonary hypertension in newborns non-invasively
• Improving accuracy of echocardiography-based PH assessment
• Enhancing model generalizability with multi-view learning

Innovation

• Multi-view VAE for PH prediction
• Captures complex latent representations
• Improves feature extraction robustness