Related Experiment Video
Updated: Jul 8, 2025
Enhancing Electrode Location Assessment in Cochlear Implantation via Computed Tomography Image Fusion
Published on: January 17, 2025
Assistive diagnostic technology for congenital heart disease based on fusion features and deep learning
Yuanlin Wang1, Xuankai Yang1, Xiaozhao Qian2
1School of Information Science and Engineering, Yunnan University, Kunming, China.
Insights
This study presents an efficient automated model for congenital heart disease (CHD) screening using heart sound analysis. The novel LCACNN model achieves high accuracy, potentially replacing manual auscultation for remote diagnostics.
Area of Science:
- Cardiology
- Biomedical Engineering
- Artificial Intelligence
Background:
- Congenital heart disease (CHD) requires early detection for effective treatment.
- Heart sound analysis aids CHD diagnosis, but automated classification models are lacking.
- Manual auscultation is the current standard but has limitations.
Purpose of the Study:
- To develop an efficient automated screening and classification model for congenital heart disease (CHD).
- To improve the accuracy and accessibility of CHD diagnosis, especially in remote areas.
Main Methods:
- A novel Locally Concatenated Fusion approach with a Convolutional Neural Network based on Coordinate Attention (LCACNN) was developed.
- Mel-frequency spectral coefficients (MFSC) and envelope features were locally fused as input.
- The model automatically analyzes feature map energy, negating the need for denoising.
Main Results:
- The LCACNN model achieved high classification accuracies: 91.78% on the PhysioNet database and 94.79% on the HS database.
- Performance significantly surpassed alternative classification models in comparative ablation experiments.
- The model demonstrated robust capability in identifying congenital heart disease.
Conclusions:
- The proposed LCACNN model offers a robust and efficient method for CHD screening and classification.
- This automated approach has the potential to substitute manual auscultation.
- It can facilitate early CHD detection, particularly benefiting patients in remote or underserved regions.
Abstract:
Introduction: Congenital heart disease (CHD) is a cardiovascular disorder caused by structural defects in the heart. Early screening holds significant importance for the effective treatment of this condition. Heart sound analysis is commonly employed to assist in the diagnosis of CHD. However, there is currently a lack of an efficient automated model for heart sound classification, which could potentially replace the manual process of auscultation. Methods: This study introduces an innovative and efficient screening and classification model, combining a locally concatenated fusion approach with a convolutional neural network based on coordinate attention (LCACNN). In this model, Mel-frequency spectral coefficients (MFSC) and envelope features are locally fused and employed as input to the LCACNN network. This model automatically analyzes feature map energy information, eliminating the need for denoising processes. Discussion: The proposed classification model in this study demonstrates a robust capability for identifying congenital heart disease, potentially substituting manual auscultation to facilitate the detection of patients in remote areas. Results: This study introduces an innovative and efficient screening and classification model, combining a locally concatenated fusion approach with a convolutional neural network based on coordinate attention (LCACNN). In this model, Mel-frequency spectral coefficients (MFSC) and envelope features are locally fused and employed as input to the LCACNN network. This model automatically analyzes feature map energy information, eliminating the need for denoising processes. To assess the performance of the classification model, comparative ablation experiments were conducted, achieving classification accuracies of 91.78% and 94.79% on the PhysioNet and HS databases, respectively. These results significantly outperformed alternative classification models.
