Related Experiment Video
Updated: Jul 30, 2025
Measurement of Tissue Oxygenation Using Near-Infrared Spectroscopy in Patients Undergoing Hemodialysis
Published on: October 2, 2020
Validation of an Automated Cardiothoracic Ratio Calculation for Hemodialysis Patients
Hsin-Hsu Chou1,2, Jin-Yi Lin3, Guan-Ting Shen3
1Department of Pediatrics, Ditmanson Medical Foundation Chia-Yi Christian Hospital, Chiayi 600566, Taiwan.
Insights
Automated cardiothoracic ratio (CTR) calculation using AlbuNet-34 demonstrated high accuracy and efficiency. This AI model offers a reliable tool for assessing cardiomegaly from chest X-rays in clinical practice.
Area of Science:
- Radiology
- Artificial Intelligence
- Nephrology
Background:
- Cardiomegaly, indicated by an elevated cardiothoracic ratio (CTR), is linked to adverse clinical outcomes.
- Accurate CTR assessment from chest X-rays (CXRs) is crucial but often subjective due to variability in operator interpretation of heart and lung margins.
Purpose of the Study:
- To develop and validate an automated method for calculating the cardiothoracic ratio (CTR) using a deep learning model.
- To compare the accuracy and efficiency of the automated CTR calculation against manual methods performed by healthcare professionals.
Main Methods:
- A U-Net variant, AlbuNet-34, was implemented to predict heart and lung margins from CXR images.
- Nephrologists provided ground truth labels for lung and heart borders.
- The model automatically calculated CTRs for patients undergoing hemodialysis.
Main Results:
- The neural network model achieved a high coefficient of determination (R² = 0.96) for CTR calculation.
- Automated CTR calculation showed significantly higher agreement with nephrologist-defined ground truth (mean difference 0.83 ± 0.87%) compared to manual calculations by nurse practitioners (mean difference 1.52 ± 1.46%).
- Automated CTR calculation reduced calculation time from 85 seconds to under 2 seconds.
Conclusions:
- Automated CTR calculation using the AlbuNet-34 model is valid and accurate.
- The AI-driven approach offers significant time savings and improved consistency, making it suitable for clinical implementation.
Abstract:
Cardiomegaly is associated with poor clinical outcomes and is assessed by routine monitoring of the cardiothoracic ratio (CTR) from chest X-rays (CXRs). Judgment of the margins of the heart and lungs is subjective and may vary between different operators.
Methods:
Patients aged > 19 years in our hemodialysis unit from March 2021 to October 2021 were enrolled. The borders of the lungs and heart on CXRs were labeled by two nephrologists as the ground truth (nephrologist-defined mask). We implemented AlbuNet-34, a U-Net variant, to predict the heart and lung margins from CXR images and to automatically calculate the CTRs.
Results:
The coefficient of determination (R2) obtained using the neural network model was 0.96, compared with an R2 of 0.90 obtained by nurse practitioners. The mean difference between the CTRs calculated by the nurse practitioners and senior nephrologists was 1.52 ± 1.46%, and that between the neural network model and the nephrologists was 0.83 ± 0.87% (p < 0.001). The mean CTR calculation duration was 85 s using the manual method and less than 2 s using the automated method (p < 0.001).
Conclusions:
Our study confirmed the validity of automated CTR calculations. By achieving high accuracy and saving time, our model can be implemented in clinical practice.
Related Concept Videos
01:24Hemodialysis II: Procedure and Complications
01:30Continuous Renal Replacement Therapy
01:25Hemodialysis III: Nursing Management
01:25Hemodialysis I: Introduction