Login
Search
Search
0 Dates
2024
2023
2022
2021
2020
2019
2018
0 Events
CPC 2018
CPC 2019
Curso de Atualização em Medicina Cardiovascular 2019
Reunião Anual Conjunta dos Grupos de Estudo de Cirurgia Cardíaca, Doenças Valvulares e Ecocardiografia da SPC
CPC 2020
CPC 2021
CPC 2022
CPC 2023
CPC 2024
0 Topics
A. Basics
B. Imaging
C. Arrhythmias and Device Therapy
D. Heart Failure
E. Coronary Artery Disease, Acute Coronary Syndromes, Acute Cardiac Care
F. Valvular, Myocardial, Pericardial, Pulmonary, Congenital Heart Disease
G. Aortic Disease, Peripheral Vascular Disease, Stroke
H. Interventional Cardiology and Cardiovascular Surgery
I. Hypertension
J. Preventive Cardiology
K. Cardiovascular Disease In Special Populations
L. Cardiovascular Pharmacology
M. Cardiovascular Nursing
N. E-Cardiology / Digital Health, Public Health, Health Economics, Research Methodology
O. Basic Science
P. Other
0 Themes
01. History of Cardiology
02. Clinical Skills
03. Imaging
04. Arrhythmias, General
05. Atrial Fibrillation
06. Supraventricular Tachycardia (non-AF)
07. Syncope and Bradycardia
08. Ventricular Arrhythmias and Sudden Cardiac Death (SCD)
09. Device Therapy
10. Chronic Heart Failure
11. Acute Heart Failure
12. Coronary Artery Disease (Chronic)
13. Acute Coronary Syndromes
14. Acute Cardiac Care
15. Valvular Heart Disease
16. Infective Endocarditis
17. Myocardial Disease
18. Pericardial Disease
19. Tumors of the Heart
20. Congenital Heart Disease and Pediatric Cardiology
21. Pulmonary Circulation, Pulmonary Embolism, Right Heart Failure
22. Aortic Disease
23. Peripheral Vascular and Cerebrovascular Disease
24. Stroke
25. Interventional Cardiology
26. Cardiovascular Surgery
27. Hypertension
28. Risk Factors and Prevention
29. Rehabilitation and Sports Cardiology
30. Cardiovascular Disease in Special Populations
31. Pharmacology and Pharmacotherapy
32. Cardiovascular Nursing
33. e-Cardiology / Digital Health
34. Public Health and Health Economics
35. Research Methodology
36. Basic Science
37. Miscellanea
0 Resources
Abstract
Slides
Vídeo
Report
CLEAR FILTERS
Machine-learning based radiomics model to predict ventricular arrythmias in patients with hypertrophic cardiomyopathy
Session:
Prémio Jovem Investigador (Clínica e Básica)
Speaker:
Miguel Marques Antunes
Congress:
CPC 2024
Topic:
P. Other
Theme:
37. Miscellanea
Subtheme:
17.3 Myocardial Disease – Diagnostic Methods
Session Type:
Prémios, Registos e Sessões Especiais
FP Number:
---
Authors:
Miguel Marques Antunes; Inês Pereira de Miranda; Vera Ferreira; Pedro Garcia Brás; José Viegas; Isabel Cardoso; Boban Thomas; Gonçalo Branco; Ricardo Pereira; Sílvia Aguiar Rosa; João Bicho Augusto; Rui Cruz Ferreira
Abstract
<p><span style="font-size:11pt"><span style="font-family:Calibri,sans-serif"><strong>Background</strong>: Patients with hypertrophic cardiomyopathy (HCM) are at a higher risk of ventricular arrythmias (VA). Late gadolinium enhancement (LGE) imaging from cardiovascular magnetic resonance (CMR) is a well-established risk factor for VAs, but the ability to predict these in various models remains sub-optimal. We hypothesized that deep quantitative phenotyping of high-dimensional data from radiomics could improve risk stratification in HCM.</span></span></p> <p><span style="font-size:11pt"><span style="font-family:Calibri,sans-serif"><strong>Objective</strong>: to assess radiomics models based on left ventricular (LV) LGE images to predict the risk of VA.</span></span></p> <p><span style="font-size:11pt"><span style="font-family:Calibri,sans-serif"><strong>Methods</strong>: Cardiac MRI images from 35 patients obtained during follow-up at a cardiomyopathy clinic were prospectively collected and reviewed. The left ventricular (LV) wall was manually segmented using the 3D Slicer 5.2.2 open access tool. We extracted several features using the PyRadiomics library (v3.1.0), including shape, first order and textural features from the LV wall. We also performed a single staged wavelet decomposition <span style="background-color:white"><span style="color:#212121">to decompose volumetric images into eight decomposed volumes of images. This led to a total of 851 parameters being extracted per patient. We performed feature relevance analysis, reducing the feature count to 623 features (removing features with a low variance of <0.01). After adjusting class weights, a Machine Learning Random Forest Classifier (ML-RFC) model was trained to predict arrhythmic events, followed by hyperparameter tuning. We then compared the performance of this model to the prognostic significance of LV </span></span><span style="background-color:white"><span style="color:black">LGE amount (</span></span>using the mean + 5 standard deviations method) and a model that included both.</span></span></p> <p><span style="font-size:11pt"><span style="font-family:Calibri,sans-serif"><strong>Results</strong>: <span style="color:black">A total of 35 HCM patients were included </span><span style="color:black">(age 54 [45-63] years, 29.4% female). All patients had LV LGE, quantified as a median 16 % [8% – 22%] of LV mass. Ventricular arrhythmias were noted in 8 (23.5%) patients. </span>The prediction models are summarized in the Figure. The ML-RFC model had a VA prediction accuracy of 85.7%, with an area under the receiver-operator characteristics curve (AUC-ROC) of 0.833. Of note, HCM patients who experienced a VA event showed features of more LV wall heterogeneity, with increased Zone Entropy (in LLL wavelet, 7.2 [interquartile range 7.0-7.3] vs 6.8 [6.6-7.0], p=0.006) and higher values of Long Run High Gray Level Emphasis (in LLH wavelet, 302 [236-337] vs 167 [88-289], p=0.040). LV LGE amount was also a predictor of VAs, with an accuracy of 71.4% and AUC-ROC of 0.667. The addition of LGE to the ML-RFC model did not result in a significant improvement of the baseline ML-RFC model (accuracy and AUC-ROC were similar, see Figure).</span></span></p> <p><span style="font-size:11pt"><span style="font-family:Calibri,sans-serif"><strong>Conclusion</strong>: We show that LV LGE images can conceal complex ultra-structural features, that can be unveiled by radiomics. These features improve prediction of ventricular arrhythmias beyond LV LGE alone, likely reflecting LV wall heterogeneity and susceptibility to arrhythmogenic mechanisms.</span></span></p>
Slides
Our mission: To reduce the burden of cardiovascular disease
Visit our site