Objective To explore the feasibility of using machine learning and audio technology and analyzing the auscultation data from arteriovenous fistula (AVF) to monitor AVF function. Methods A total of 50 patients with AVF stenosis which required percutaneous transluminal angioplasty (PTA) were recruited in this study. Digital sound of the AVF shunt was recorded using a wireless electronic stethoscope before (pre-PTA) and after PTA (post-PTA). The audio files were subsequently converted to spectrograms, and the acoustic signatures were extracted. More than 30 classification learners in MATLAB software were used to analyze the extracted acoustic features, and models to evaluate the AVF function were constructed. Finally, the performance of these models was compared. Results One hundred audio files were obtained from the 50 recruited patients and were pooled for the study. In the spectrogram comparison of AVF before PTA and after PTA, the spectrogram before PTA showed a larger high-frequency amplitude. There was a significant difference between the files before and after PTA (Z=-4.721, P＜0.001). The highest and lowest frequencies in a cardiac cycle were significant different before and after PTA, and the difference was greater before PTA than after PTA (Z=-6.169, P＜0.001). Among the models established by more than 30 kinds of classification learners, the model established by efficient linear support vector machine (SVM) and coarse Gaussian SVM had the best performance, with 81.11% accuracy after 5-fold cross validation. The test accuracy of the model constructed by quadratic discrimination, cubic kNN, medium neural network, and double-layer neural network can reach up to 90%. Conclusions Spectrum-based machine learning models can predict the saliency narrowness of AVF, so it is feasible to use it for the monitoring of AVF function. The model constructed by efficient linear SVM and rough Gaussian SVM had the best performance in this feasibility study.
WANG Fan-Li
,
YANG Yan-Li
,
XU Yuan-Kai
,
RUAN Lin
,
LI Wen
,
LIU Yong-Liang
,
SUN Li-Jun
,
LEI Ying
,
LIU Xiao-Ming
,
ZHAO Pei-Nan
,
ZHANG Li-Hong
. Feasibility of arteriovenous fistula function monitoring based on machine learning and audio technology[J]. Chinese Journal of Blood Purification, 2024
, 23(09)
: 701
-705
.
DOI: 10.3969/j.issn.1671-4091.2024.09.015
[1] 金其庄, 王玉柱, 叶朝阳, 等. 中国血液透析用血管通路专家共识(第2版)[J]. 中国血液净化, 2019, 18(06): 365-381.
[2] Lok CE, Huber TS, Lee T, et al. KDOQI clinical practice guideline for vascular access: 2019 update[J]. Am J Kidney Dis, 2020, 75(4): S1-S164.
[3] Schmidli J, Widmer MK, Basile C, et al. Editor's choice-vascular access: 2018 clinical practice guidelines of the European Society for Vascular Surgery (ESVS)[J]. Eur J Vasc Endovasc Surg, 2018, 55(6): 757-818.
[4] Tsuboi M, Suzuki H, Kawai H, et al. Vascular sound visualization system is useful for monitoring and surveillance of vascular access[J]. J Vasc Access, 2022, 23(3): 390-397.
[5] Grochowina M, Leniowska L, Gala-B??dzińska A. The prototype device for non-invasive diagnosis of arteriovenous fistula condition using machine learning methods[J]. Sci Rep, 2020, 10(1): 16387.
[6] Panda B, Mandal S, Majerus SJA. Flexible, skin coupled microphone array for point of care vascular access monitoring[J]. IEEE Trans Biomed Circuits Syst, 2019, 13(6): 1494-1505.
[7] Bardou D, Zhang K, Ahmad SM. Lung sounds classification using convolutional neural networks. Artif Intell Med 2018;88:58-69
[8] Messner E, Fediuk M, Swatek P, Scheidl S, Smolle-Jüttner FM, Olschewski H, et al. Multi-channel lung sound classification with convolutional recurrent neural networks. Comput Biol Med 2020;122:103831
[9] Salman L, Beathard G. Interventional nephrology: physical examination as a tool for surveillance for the hemodialysis arteriovenous access[J]. Clin J Am Soc Nephrol, 2013, 8(7): 1220-1227.
[10]Noor AM. Non-imaging acoustical properties in monitoring arteriovenous hemodialysis access. A review[J]. ?J Eng Technol Sci, 2015, 47(6): 658-673. DOI: https://doi.org/10.5614/j.eng.technol.sci.2015.47.6.6.
[11] Munguía MM, Vasquez P, Mondersson B, et al. Characterisation of Arteriovenous Fistula’s Sound Recordings using Principal Component Analysis[C]. 2009 Annu Int Conf IEEE Eng Med Biol Soc, Minneapolis, MN, USA: IEEE, 2009: 5661-5664.
[12] Munguía MM, Mandersson B. Analysis of the vascular sounds of the arteriovenous fistula's anastomosis[C]. 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Boston, MA: IEEE, 2011: 3784-3787.
[13] Gram M, Olesen J T, Riis H C, et al. Stenosis detection algorithm for screening of arteriovenous fistulae[C]. 15th Nordic-Baltic Conference on Biomedical Engineering and Medical Physics. Denmark: NBC 2011, 2011: 241-244.
[14] Obando PV, Mandersson B. Frequency tracking of resonant-like sounds from audio recordings of arterio-venous fistula stenosis[C]. 2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops. Philadelphia, PA: IEEE, 2012: 771-773.
[15] Todo A, Kadonaka T, Yoshioka M, et al. Frequency analysis of shunt sounds in the arteriovenous fistula on hemodialysis patients[C]. The 6th International Conference on Soft Computing and Intelligent Systems, and The 13th International Symposium on Advanced Intelligence Systems. Kobe: IEEE, 2012: 1113-1118.
[16] Wang YN, Chan CY, Chou SJ. The detection of arteriovenous fistula stenosis for hemodialysis based on wavelet transform[J]. Int J Adv Comput Sci Appl, 2011, 1(1): 16-22.
[17] Mansy HA, Hoxie SJ, Patel NH, et al. Computerised analysis of auscultatory sounds associated with vascular patency of haemodialysis access[J]. Med Biol Eng Comput, 2005, 43(1): 56-62.
[18] Munguía M, Vásquez P, Mattsson E, et al. Acoustical detection of venous stenosis in hemodialysis patients using principal component analysis[C]. 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology. Buenos Aires: IEEE, 2010: 3654-3657.
[19] Fredberg JJ. Origin and character of vascular murmurs: Model studies[J]. J Acoust Soc Am, 1977, 61(4): 1077-1085.
[20] Wang JZ, Tie B, Welkowitz W, et al. Modeling sound generation in stenosed coronary arteries[J]. IEEE Trans Biomed Eng, 1990, 37(11): 1087-1094.
[21] Akay Y, Akay M, Welkowitz W, et al. Dynamics of the sounds caused by partially occluded femoral arteries in dogs[J]. Ann Biomed Eng, 1994, 22(5): 493-500.
[22] Sato T, Tsuji K, Kawashima N, et al. Evaluation of blood access dysfunction based on a wavelet transform analysis of shunt murmurs[J]. J Artif Organs, 2006, 9(2): 97-104.
[23]Park JH, Park I, Han K, et al. Feasibility of Deep Learning-Based Analysis of Auscultation for Screening Significant Stenosis of Native Arteriovenous Fistula for Hemodialysis Requiring Angioplasty[J]. Korean J Radiol. 2022;23(10):949-958.
[24] Ota K, Nishiura Y, Ishihara S, et al. Evaluation of hemodialysis arteriovenous bruit by deep learning[J]. Sensors, 2020, 20(17): 4852.
[25]Grochowina M, Leniowska L, Gala-B??dzińska A. The prototype device for non-invasive diagnosis of arteriovenous fistula condition using machine learning methods[J].?Sci Rep. 2020;10(1):16387. Published 2020 Oct 2. doi:10.1038/s41598-020-72336-5.
