Prediction of Arteriovenous Access Dysfunction by Mel Spectrogram-based Deep Learning Model
International Journal of Medical Sciences · 2024 · 21 (12) · 2252-2260
DOI: 10.7150/ijms.98421 ↗At a glance
Journal article in International Journal of Medical Sciences, 2024. Written by Tung-Ling Chung and 8 co-authors. Institution: Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan. Major MEDLINE topics: Deep Learning, Renal Dialysis, Arteriovenous Shunt, Surgical. Cited 13 times so far (Crossref). The full text is freely available in PubMed Central.
Abstract (original language)
Background
The early detection of arteriovenous (AV) access dysfunction is crucial for maintaining the patency of vascular access. This study aimed to use deep learning to predict AV access malfunction necessitating further vascular management.
Methods
This prospective cohort study enrolled prevalent hemodialysis (HD) patients with an AV fistula or AV graft from a single HD center. Their AV access bruit sounds were recorded weekly using an electronic stethoscope from three different sites (arterial needle site, venous needle site, and the midpoint between the arterial and venous needle sites) before HD sessions. The audio signals were converted to Mel spectrograms using Fourier transformation and utilized to develop deep learning models. Three deep learning models, (1) Convolutional Neural Network (CNN), (2) Convolutional Recurrent Neural Network (CRNN), and (3) Vision Transformers-Gate Recurrent Unit (ViT-GRU), were trained and compared to predict the likelihood of dysfunctional AV access.
Results
Total 437 audio recordings were obtained from 84 patients. The CNN model outperformed the other models in the test set, with an F1 score of 0.7037 and area under the receiver operating characteristic curve (AUROC) of 0.7112. The Vit-GRU model had high performance in out-of-fold predictions, with an F1 score of 0.7131 and AUROC of 0.7745, but low generalization ability in the test set, with an F1 score of 0.5225 and AUROC of 0.5977.
Conclusions
The CNN model based on Mel spectrograms could predict malfunctioning AV access requiring vascular intervention within 10 days. This approach could serve as a useful screening tool for high-risk AV access.
Subjects
MeSH terms (MEDLINE)
Author keywords
Authors and institutions
- Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan.
- Division of Nephrology, Department of Internal Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan.
- Graduate Institute of Clinical Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan.
- Department of Internal Medicine, Kaohsiung Municipal Siaogang Hospital, Kaohsiung Medical University Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan.
- Division of Cardiology, Department of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan.
- Division of Nephrology, Department of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan.
Cite
APA
Chung, T. L., Liu, Y. H., Wu, P. Y., Huang, J. C., Tsai, Y. C., Wang, Y. C., Pan, S. P., Hsu, Y. L., & Chen, S. C. (2024). Prediction of Arteriovenous Access Dysfunction by Mel Spectrogram-based Deep Learning Model. International Journal of Medical Sciences, 21(12), 2252-2260. https://doi.org/10.7150/ijms.98421
Vancouver
Chung TL, Liu YH, Wu PY, Huang JC, Tsai YC, Wang YC, et al. Prediction of Arteriovenous Access Dysfunction by Mel Spectrogram-based Deep Learning Model. International Journal of Medical Sciences. 2024;21(12):2252-2260. doi:10.7150/ijms.98421
BibTeX
@article{Chung2024,
author = {Chung, Tung-Ling and Liu, Yi-Hsueh and Wu, Pei-Yu and Huang, Jiun-Chi and Tsai, Yi-Chun and Wang, Yu-Chen and Pan, Shan-Pin and Hsu, Ya-Ling and Chen, Szu-Chia},
title = {Prediction of Arteriovenous Access Dysfunction by Mel Spectrogram-based Deep Learning Model},
journal = {International Journal of Medical Sciences},
year = {2024},
volume = {21},
number = {12},
pages = {2252--2260},
doi = {10.7150/ijms.98421}
}
EUVIMED