Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CardioMOD-Net: A Modal Decomposition-Neural Network Framework for Diagnosis and Prognosis of HFpEF from Echocardiography Cine Loops

Andrés Bell-Navas, Jesús Garicano-Mena, Antonella Ausiello, Soledad Le Clainche, María Villalba-Orero, Enrique Lara-Pezzi

TL;DR

HFpEF presents a heterogeneous, comorbidity-driven progression, challenging early diagnosis and time-to-event prognosis. CardioMOD-Net unifies multiclass phenotyping and continuous onset prediction from standard PLAX cine loops by extracting spatiotemporal features with Higher Order Dynamic Mode Decomposition and using a shared latent representation to power both a Vision Transformer classifier and a regression head for time-to-HFpEF onset, achieving $65\%$ overall accuracy and a RMSE of $21.72$ weeks in a small preclinical dataset. The method demonstrates that meaningful diagnostic and prognostic information can be derived from raw cine data without Doppler indices or clinical metadata, and that integrated, real-time inference is feasible. This framework lays groundwork for linking comorbidity-specific myocardial dynamics to clinical risk in HFpEF and may inform translational preclinical and early-human studies.

Abstract

Introduction: Heart failure with preserved ejection fraction (HFpEF) arises from diverse comorbidities and progresses through prolonged subclinical stages, making early diagnosis and prognosis difficult. Current echocardiography-based Artificial Intelligence (AI) models focus primarily on binary HFpEF detection in humans and do not provide comorbidity-specific phenotyping or temporal estimates of disease progression towards decompensation. We aimed to develop a unified AI framework, CardioMOD-Net, to perform multiclass diagnosis and continuous prediction of HFpEF onset directly from standard echocardiography cine loops in preclinical models. Methods: Mouse echocardiography videos from four groups were used: control (CTL), hyperglycaemic (HG), obesity (OB), and systemic arterial hypertension (SAH). Two-dimensional parasternal long-axis cine loops were decomposed using Higher Order Dynamic Mode Decomposition (HODMD) to extract temporal features for downstream analysis. A shared latent representation supported Vision Transformers, one for a classifier for diagnosis and another for a regression module for predicting the age at HFpEF onset. Results: Overall diagnostic accuracy across the four groups was 65%, with all classes exceeding 50% accuracy. Misclassifications primarily reflected early-stage overlap between OB or SAH and CTL. The prognostic module achieved a root-mean-square error of 21.72 weeks for time-to-HFpEF prediction, with OB and SAH showing the most accurate estimates. Predicted HFpEF onset closely matched true distributions in all groups. Discussion: This unified framework demonstrates that multiclass phenotyping and continuous HFpEF onset prediction can be obtained from a single cine loop, even under small-data conditions. The approach offers a foundation for integrating diagnostic and prognostic modelling in preclinical HFpEF research.

CardioMOD-Net: A Modal Decomposition-Neural Network Framework for Diagnosis and Prognosis of HFpEF from Echocardiography Cine Loops

TL;DR

HFpEF presents a heterogeneous, comorbidity-driven progression, challenging early diagnosis and time-to-event prognosis. CardioMOD-Net unifies multiclass phenotyping and continuous onset prediction from standard PLAX cine loops by extracting spatiotemporal features with Higher Order Dynamic Mode Decomposition and using a shared latent representation to power both a Vision Transformer classifier and a regression head for time-to-HFpEF onset, achieving overall accuracy and a RMSE of weeks in a small preclinical dataset. The method demonstrates that meaningful diagnostic and prognostic information can be derived from raw cine data without Doppler indices or clinical metadata, and that integrated, real-time inference is feasible. This framework lays groundwork for linking comorbidity-specific myocardial dynamics to clinical risk in HFpEF and may inform translational preclinical and early-human studies.

Abstract

Introduction: Heart failure with preserved ejection fraction (HFpEF) arises from diverse comorbidities and progresses through prolonged subclinical stages, making early diagnosis and prognosis difficult. Current echocardiography-based Artificial Intelligence (AI) models focus primarily on binary HFpEF detection in humans and do not provide comorbidity-specific phenotyping or temporal estimates of disease progression towards decompensation. We aimed to develop a unified AI framework, CardioMOD-Net, to perform multiclass diagnosis and continuous prediction of HFpEF onset directly from standard echocardiography cine loops in preclinical models. Methods: Mouse echocardiography videos from four groups were used: control (CTL), hyperglycaemic (HG), obesity (OB), and systemic arterial hypertension (SAH). Two-dimensional parasternal long-axis cine loops were decomposed using Higher Order Dynamic Mode Decomposition (HODMD) to extract temporal features for downstream analysis. A shared latent representation supported Vision Transformers, one for a classifier for diagnosis and another for a regression module for predicting the age at HFpEF onset. Results: Overall diagnostic accuracy across the four groups was 65%, with all classes exceeding 50% accuracy. Misclassifications primarily reflected early-stage overlap between OB or SAH and CTL. The prognostic module achieved a root-mean-square error of 21.72 weeks for time-to-HFpEF prediction, with OB and SAH showing the most accurate estimates. Predicted HFpEF onset closely matched true distributions in all groups. Discussion: This unified framework demonstrates that multiclass phenotyping and continuous HFpEF onset prediction can be obtained from a single cine loop, even under small-data conditions. The approach offers a foundation for integrating diagnostic and prognostic modelling in preclinical HFpEF research.
Paper Structure (13 sections, 1 figure)

This paper contains 13 sections, 1 figure.

Table of Contents

  1. Introduction
  2. Methods
  3. Data acquisition and experimental groups
  4. Dataset structure and train–test separation
  5. AI framework and feature extraction
  6. Diagnosis module
  7. Prognosis module
  8. Real-time inference
  9. Results
  10. Diagnosis performance
  11. Prognostic performance
  12. Integrated task performance
  13. Discussion

Figures (1)

  • Figure 1: Diagnosis and prognosis results obtained with the CardioMOD-Net tool (Cardiovascular Modal Decomposition - Neural Network). A, Schematic depicting the workflow used by this AI-based tool for heart disease recognition. Top, training of the tool for diagnosis and prognosis; bottom, diagnosis and prognosis performed by the tool after it is trained. B, Summary of the main characteristics of echocardiography images in our database. C, Diagnosis results obtained with the tool using the test sequences. D, Confusion matrix obtained with the developed tool using test sequences broken down into age intervals. Top, global confusion matrix; bottom, confusion matrices by age intervals. E, Prognosis results obtained with the tool using the test sequences. CTL, control; HG, hyperglycaemic; OB, obesity; SAH, systemic arterial hypertension; RMSE, Root Mean Squared Error.