LightCardiacNet: light-weight deep ensemble network with attention mechanism for cardiac sound classification
Cardiovascular diseases (CVDs) account for about 32% of global deaths. While digital stethoscopes can record heart sounds, expert analysis is often lacking. To address this, we propose LightCardiacNet, an interpretable, lightweight ensemble neural network using Bi-Directional Gated Recurrent Units (...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2024-12-01
|
| Series: | Systems Science & Control Engineering |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/21642583.2024.2420912 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846118821942788096 |
|---|---|
| author | Suma K. V. Deepali B. Koppad Dharini Raghavan Manjunath P. R. |
| author_facet | Suma K. V. Deepali B. Koppad Dharini Raghavan Manjunath P. R. |
| author_sort | Suma K. V. |
| collection | DOAJ |
| description | Cardiovascular diseases (CVDs) account for about 32% of global deaths. While digital stethoscopes can record heart sounds, expert analysis is often lacking. To address this, we propose LightCardiacNet, an interpretable, lightweight ensemble neural network using Bi-Directional Gated Recurrent Units (Bi-GRU). It is trained on the PASCAL Heart Challenge and CirCor DigiScope datasets. Static network pruning enhances model sparsity for real-time deployment. We employ various data augmentation techniques to improve resilience to background noise. An ensemble of the two networks is constructed by employing a weighted average approach that combines the two light-weight attention Bi-GRU networks trained on different datasets, which outperforms several state-of-the-art networks achieving an accuracy of 99.8%, specificity of 99.6%, sensitivity of 95.2%, ROC-AUC of 0.974 and inference time of 17 ms on the PASCAL dataset, accuracy of 98.5%, specificity of 95.1%, sensitivity of 90.9%, ROC-AUC of 0.961 and inference time of 18 ms on the CirCor dataset, and an accuracy of 96.21%, sensitivity of 92.78%, specificity of 93.16%, ROC-AUC of 0.913 and inference time of 17.5 ms on real-world data. We adopt the SHAP algorithm to incorporate model interpretability and provide insights to make it clinically explainable and useful to healthcare professionals. |
| format | Article |
| id | doaj-art-841ffe9c6c5a4c54a29a1c4474912e4c |
| institution | Kabale University |
| issn | 2164-2583 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Systems Science & Control Engineering |
| spelling | doaj-art-841ffe9c6c5a4c54a29a1c4474912e4c2024-12-17T09:06:12ZengTaylor & Francis GroupSystems Science & Control Engineering2164-25832024-12-0112110.1080/21642583.2024.2420912LightCardiacNet: light-weight deep ensemble network with attention mechanism for cardiac sound classificationSuma K. V.0Deepali B. Koppad1Dharini Raghavan2Manjunath P. R.3Department of ECE, Ramaiah Institute of Technology, Bangalore, IndiaDepartment of ECE, Ramaiah Institute of Technology, Bangalore, IndiaDepartment of ECE, Ramaiah Institute of Technology, Bangalore, IndiaDepartment of Endocrinology, Ramaiah Medical College, Bangalore, IndiaCardiovascular diseases (CVDs) account for about 32% of global deaths. While digital stethoscopes can record heart sounds, expert analysis is often lacking. To address this, we propose LightCardiacNet, an interpretable, lightweight ensemble neural network using Bi-Directional Gated Recurrent Units (Bi-GRU). It is trained on the PASCAL Heart Challenge and CirCor DigiScope datasets. Static network pruning enhances model sparsity for real-time deployment. We employ various data augmentation techniques to improve resilience to background noise. An ensemble of the two networks is constructed by employing a weighted average approach that combines the two light-weight attention Bi-GRU networks trained on different datasets, which outperforms several state-of-the-art networks achieving an accuracy of 99.8%, specificity of 99.6%, sensitivity of 95.2%, ROC-AUC of 0.974 and inference time of 17 ms on the PASCAL dataset, accuracy of 98.5%, specificity of 95.1%, sensitivity of 90.9%, ROC-AUC of 0.961 and inference time of 18 ms on the CirCor dataset, and an accuracy of 96.21%, sensitivity of 92.78%, specificity of 93.16%, ROC-AUC of 0.913 and inference time of 17.5 ms on real-world data. We adopt the SHAP algorithm to incorporate model interpretability and provide insights to make it clinically explainable and useful to healthcare professionals.https://www.tandfonline.com/doi/10.1080/21642583.2024.2420912Cardiovascular diseasesneural networkssparsitygated recurrent unitslong short-term memory networksensemble learning |
| spellingShingle | Suma K. V. Deepali B. Koppad Dharini Raghavan Manjunath P. R. LightCardiacNet: light-weight deep ensemble network with attention mechanism for cardiac sound classification Systems Science & Control Engineering Cardiovascular diseases neural networks sparsity gated recurrent units long short-term memory networks ensemble learning |
| title | LightCardiacNet: light-weight deep ensemble network with attention mechanism for cardiac sound classification |
| title_full | LightCardiacNet: light-weight deep ensemble network with attention mechanism for cardiac sound classification |
| title_fullStr | LightCardiacNet: light-weight deep ensemble network with attention mechanism for cardiac sound classification |
| title_full_unstemmed | LightCardiacNet: light-weight deep ensemble network with attention mechanism for cardiac sound classification |
| title_short | LightCardiacNet: light-weight deep ensemble network with attention mechanism for cardiac sound classification |
| title_sort | lightcardiacnet light weight deep ensemble network with attention mechanism for cardiac sound classification |
| topic | Cardiovascular diseases neural networks sparsity gated recurrent units long short-term memory networks ensemble learning |
| url | https://www.tandfonline.com/doi/10.1080/21642583.2024.2420912 |
| work_keys_str_mv | AT sumakv lightcardiacnetlightweightdeepensemblenetworkwithattentionmechanismforcardiacsoundclassification AT deepalibkoppad lightcardiacnetlightweightdeepensemblenetworkwithattentionmechanismforcardiacsoundclassification AT dhariniraghavan lightcardiacnetlightweightdeepensemblenetworkwithattentionmechanismforcardiacsoundclassification AT manjunathpr lightcardiacnetlightweightdeepensemblenetworkwithattentionmechanismforcardiacsoundclassification |