A Multimodular AI Algorithm for Automated Assessment of Left Ventricular Function in Ischemic Heart Disease - Ejection Fraction, Wall Motion, and Regional Myocardial Segmentation

Q: What clinical problem does the proposed AI algorithm address?

It addresses experience-dependent interpretation of 2D-TTE for assessing left ventricular function, specifically LVEF calculation and WMA detection in ischemic heart disease.

Q: How does the algorithm compute LVEF and detect wall motion abnormalities?

It uses two modules: a static frame segmentation combined with dynamic tracking for LVEF via a hybrid Simpson’s method, and a wall motion module using YOLOv8 for segmentation plus ResNet50 for performance assessment.

Q: How were the model results evaluated against cardiologist measurements?

LVEF performance was assessed using Bland-Altman analysis and regression correlation with cardiologist measurements, while wall motion performance was evaluated using accuracy and related metrics.

Sep 28, 2026

Uploaded by nayy☆

Ischemic heart disease impairs left ventricular pumping by reducing ejection fraction (LVEF) and causing wall motion abnormality (WMA), which in practice is assessed from 2D transthoracic echocardiography (2D-TTE) images through experience-dependent visual interpretation. A diagnostic-accuracy study develops a multimodule artificial intelligence workflow to automate LVEF calculation and WMA detection from apical 2- and 4-chamber views. In 600 retrospectively included adult patients, the LVEF module shows strong agreement with cardiologist measurements, while the wall motion module combines YOLOv8 segmentation and ResNet50 classification, providing standardized color-coded outputs.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages13

Reading Time33 min

AccessFree to download

Background

Aims

Study Design

Methods

Results

Conclusion

Corresponding Author

Original Article 25 Content of this journal is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Balkan Med J 2026;43:25-37 A Multimodular AI Algorithm for Automated Assessment of Left Ventricular Function in Ischemic Heart Disease: Ejection Fraction, Wall Motion, and Regional Myocardial Segmentation Sidem Gül1 , Reşit Taşdemir2 , Beyza Açıkgöz2 , Hakan Duman3 , Hamza Hodzic4 , Sena Köker5 , Nazlı Erdemir6 , Mehmet Kıvrak7 1Department of Medicine, Recep Tayyip Erdoğan University Faculty of Medicine, Rize, Türkiye 2Department of Software Engineering, Malatya Turgut Özal University Faculty of Engineering and Natural Sciences, Malatya, Türkiye 3Department of Cardiology, Recep Tayyip Erdoğan University Faculty of Medicine, Rize, Türkiye 4Turkish Red Crescent Blood Donation Center, Kastamonu, Türkiye 5Department of Medicine, İstanbul University Faculty of Medicine, İstanbul, Türkiye 6Department of Biomedical Engineering, Kocaeli University Faculty of Technology, Kocaeli, Türkiye 7Department of Biostatistics and Medical Infıormatics, Recep Tayyip Erdoğan University Faculty of Medicine, Rize, Türkiye Background: Ischemic heart damage reduces the pumping efficiency of the heart by affecting the left ventricular ejection fraction (LVEF) and causing wall motion abnormality (WMA) . In daily clinical practice, these parameters are interpreted by physicians using two dimensional transthoracic echocardiography (2D-TTE) . Because 2D-TTE reports rely on visual evaluations, they are subject to experience-based limitations and exhibit low reproducibility. Aims: To develop an artificial intelligence algorithm composed of two modules that enable automatic LVEF calculation and WMA detection for analyzing 2D-TTE images. Study Design: Diagnostic accuracy study. Methods: A total of 600 adult patients were retrospectively included. The model combined static frame segmentation with dynamic tracking using a hybrid Simpson’s method applied to apical 2- and 4-chamber views. Model performance was assessed against cardiologist measurements using Bland-Altman analysis. The YOLOv8 and ResNet50 models were employed for the wall motion module. Performance metrics, including accuracy, precision, F1 score, and area under the curve, were evaluated. Results: In the Bland-Altman analysis, the mean bias between the LVEF module and cardiologist measurements was-4, with limits of agreement ranging from -15 to -3. Regression analysis demonstrated a strong correlation between the LVEF module and cardiologist measurements (r = 0 .71, p < 0 .001) . In the wall motion module, the YOLOv8 segmentation model exhibited high accuracy, while ResNet50 achieved superior performance with an accuracy of 95% . The algorithm’s color coding contributed to standardized interpretation among operators, enhancing consistency. Conclusion: This is the first study to integrate automated EF calculation and WMA detection within a single workflow. SafeHeart offers accurate, reproducible, and rapid analysis, with the potential to support routine echocardiography practice. Color-coded region segmentation can facilitate more standardized and reliable results. Corresponding author: Sidem Gül, Department of Medicine, Recep Tayyip Erdoğan University Faculty of Medicine, Rize, Türkiye e-mail: [sidem_gul21@erdogan.edu.tr](sidem_gul21@erdogan.edu.tr) Received: August 18, 2025 Accepted: October 12, 2025 Available Online Date: January 02, 2026 • DOI: 10.4274/balkanmedj. galenos.2025.2025-8-160 Available [at www.balkanmedicaljournal.org](at www.balkanmedicaljournal.org) ORCID iDs of the authors: S. G. 0009-0000-0379-1002; R.T. 0009-0008-9369-2910; B.A. 0009-0007-6770-0588; H. D. 0000-0002-1441-7320; H. H. 0009-0007-6940-9006; S. K. 0009-0006-2628-5181; N. E. 0009-0005-8517-5102; M. K. 0000-0002-2405-8552. Cite this article as: Gül S, Taşdemir R, Açıkgöz B, et al. A Multimodular AI Algorithm for Automated A

FAQs

What clinical problem does the proposed AI algorithm address?

It addresses experience-dependent interpretation of 2D-TTE for assessing left ventricular function, specifically LVEF calculation and WMA detection in ischemic heart disease.

How does the algorithm compute LVEF and detect wall motion abnormalities?

It uses two modules: a static frame segmentation combined with dynamic tracking for LVEF via a hybrid Simpson’s method, and a wall motion module using YOLOv8 for segmentation plus ResNet50 for performance assessment.

How were the model results evaluated against cardiologist measurements?

LVEF performance was assessed using Bland-Altman analysis and regression correlation with cardiologist measurements, while wall motion performance was evaluated using accuracy and related metrics.

Sep 28, 2026

Uploaded by nayy☆