An Explainable AI-Integrated Diagnostic System for Voice Analysis in Heart Failure Patients

Mikolaj Najda, Milosz Dudek, Olgierd Unold, Tomasz Jadczyk, Krzysztof Swierz, Grzegorz Swiatek, Daria Hemmerling
Proceedings of The First AAAI Bridge Program on AI for Medicine and Healthcare, PMLR 281:56-62, 2025.

Abstract

Integrating Explainable Artificial Intelligence to analyse voice characteristics is an essential topic for future research. We explore the utility of tree-based machine learning models, including Random Forest, XGBoost, and LightGBM, in distinguishing between two groups: 100 participants with heart failure and 100 healthy controls. The acoustic features extracted from sustained vowel recordings are used to differentiate between the two groups. The evaluation shows that the Random Forest model performs better, especially with the vowel /i/, achieving Accuracy, Precision, Recall, and F1 score over 0.80. We investigate the interpretability of these models usingSHapleyAdditiveexPlanationsvalues,whichrevealthe essential acoustic features that influence model predictions and provide insights into their clinical relevance. This research highlights the potential of interpretable vocal biomarkers in remote monitoring and diagnosing heart failure.

Cite this Paper

BibTeX
@InProceedings{pmlr-v281-najda25a, title = {An Explainable AI-Integrated Diagnostic System for Voice Analysis in Heart Failure Patients}, author = {Najda, Mikolaj and Dudek, Milosz and Unold, Olgierd and Jadczyk, Tomasz and Swierz, Krzysztof and Swiatek, Grzegorz and Hemmerling, Daria}, booktitle = {Proceedings of The First AAAI Bridge Program on AI for Medicine and Healthcare}, pages = {56--62}, year = {2025}, editor = {Wu, Junde and Zhu, Jiayuan and Xu, Min and Jin, Yueming}, volume = {281}, series = {Proceedings of Machine Learning Research}, month = {25 Feb}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v281/main/assets/najda25a/najda25a.pdf}, url = {https://proceedings.mlr.press/v281/najda25a.html}, abstract = {Integrating Explainable Artificial Intelligence to analyse voice characteristics is an essential topic for future research. We explore the utility of tree-based machine learning models, including Random Forest, XGBoost, and LightGBM, in distinguishing between two groups: 100 participants with heart failure and 100 healthy controls. The acoustic features extracted from sustained vowel recordings are used to differentiate between the two groups. The evaluation shows that the Random Forest model performs better, especially with the vowel /i/, achieving Accuracy, Precision, Recall, and F1 score over 0.80. We investigate the interpretability of these models usingSHapleyAdditiveexPlanationsvalues,whichrevealthe essential acoustic features that influence model predictions and provide insights into their clinical relevance. This research highlights the potential of interpretable vocal biomarkers in remote monitoring and diagnosing heart failure.} }
Endnote
%0 Conference Paper %T An Explainable AI-Integrated Diagnostic System for Voice Analysis in Heart Failure Patients %A Mikolaj Najda %A Milosz Dudek %A Olgierd Unold %A Tomasz Jadczyk %A Krzysztof Swierz %A Grzegorz Swiatek %A Daria Hemmerling %B Proceedings of The First AAAI Bridge Program on AI for Medicine and Healthcare %C Proceedings of Machine Learning Research %D 2025 %E Junde Wu %E Jiayuan Zhu %E Min Xu %E Yueming Jin %F pmlr-v281-najda25a %I PMLR %P 56--62 %U https://proceedings.mlr.press/v281/najda25a.html %V 281 %X Integrating Explainable Artificial Intelligence to analyse voice characteristics is an essential topic for future research. We explore the utility of tree-based machine learning models, including Random Forest, XGBoost, and LightGBM, in distinguishing between two groups: 100 participants with heart failure and 100 healthy controls. The acoustic features extracted from sustained vowel recordings are used to differentiate between the two groups. The evaluation shows that the Random Forest model performs better, especially with the vowel /i/, achieving Accuracy, Precision, Recall, and F1 score over 0.80. We investigate the interpretability of these models usingSHapleyAdditiveexPlanationsvalues,whichrevealthe essential acoustic features that influence model predictions and provide insights into their clinical relevance. This research highlights the potential of interpretable vocal biomarkers in remote monitoring and diagnosing heart failure.
APA
Najda, M., Dudek, M., Unold, O., Jadczyk, T., Swierz, K., Swiatek, G. & Hemmerling, D.. (2025). An Explainable AI-Integrated Diagnostic System for Voice Analysis in Heart Failure Patients. Proceedings of The First AAAI Bridge Program on AI for Medicine and Healthcare, in Proceedings of Machine Learning Research 281:56-62 Available from https://proceedings.mlr.press/v281/najda25a.html.

Related Material