Estimation of Stroke Volume Variance from Arterial Blood Pressure: Using a 1-D Convolutional Neural Network
Abstract
:1. Introduction
2. Materials and Methods
3. Results
4. Discussion
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Morgan, B.C.; Crawford, E.W.; Hornbein, T.F.; Martin, W.E.; Guntheroth, W.G. Hemodynamic effects of changes in arterial carbon dioxide tension during intermittent positive pressure ventilation. Anesthesiology 1967, 28, 866–873. [Google Scholar] [CrossRef]
- Teboul, J.L.; Monnet, X.; Chemla, D.; Michard, F. Arterial Pulse Pressure Variation with Mechanical Ventilation. Am. J. Respir. Crit. Care Med. 2019, 199, 22–31. [Google Scholar] [CrossRef]
- Vieillard-Baron, A.; Chergui, K.; Augarde, R.; Prin, S.; Page, B.; Beauchet, A.; Jardin, F. Cyclic changes in arterial pulse during respiratory support revisited by Doppler echocardiography. Am. J. Respir. Crit. Care Med. 2003, 168, 671–676. [Google Scholar] [CrossRef]
- Rex, S.; Brose, S.; Metzelder, S.; Huneke, R.; Schalte, G.; Autschbach, R.; Rossaint, R.; Buhre, W. Prediction of fluid responsiveness in patients during cardiac surgery. Br. J. Anaesth. 2004, 93, 782–788. [Google Scholar] [CrossRef] [Green Version]
- Berkenstadt, H.; Margalit, N.; Hadani, M.; Friedman, Z.; Segal, E.; Villa, Y.; Perel, A. Stroke volume variation as a predictor of fluid responsiveness in patients undergoing brain surgery. Anesth. Analg. 2001, 92, 984–989. [Google Scholar] [CrossRef]
- Li, Q.; Cai, W.; Wang, X.; Zhou, Y.; Feng, D.D.; Chen, M. Medical image classification with convolutional neural network. In Proceedings of the 13th International Conference on Control Automation Robotics & Vision (ICARCV), Singapore, 10–12 December 2014; pp. 844–848. [Google Scholar]
- Marx, G.; Cope, T.; McCrossan, L.; Swaraj, S.; Cowan, C.; Mostafa, S.M.; Wenstone, R.; Leuwer, M. Assessing fluid responsiveness by stroke volume variation in mechanically ventilated patients with severe sepsis. Eur. J. Anaesthesiol. 2004, 21, 132–138. [Google Scholar] [CrossRef]
- Reuter, D.A.; Kirchner, A.; Felbinger, T.W.; Weis, F.C.; Kilger, E.; Lamm, P.; Goetz, A.E. Usefulness of left ventricular stroke volume variation to assess fluid responsiveness in patients with reduced cardiac function. Crit. Care Med. 2003, 31, 1399–1404. [Google Scholar] [CrossRef] [PubMed]
- Park, M.; Han, S.; Kim, G.S.; Gwak, M.S. Evaluation of new calibrated pulse-wave analysis (VolumeViewTM/EV1000TM) for cardiac output monitoring undergoing living donor liver transplantation. PLoS ONE 2016, 11, e0164521. [Google Scholar] [CrossRef]
- Wiesenack, C.; Prasser, C.; Rodig, G.; Keyl, C. Stroke volume variation as an indicator of fluid responsiveness using pulse contour analysis in mechanically ventilated patients. Anesth. Analg. 2003, 96, 1254–1257. [Google Scholar] [CrossRef]
- Buhre, W.; Rossaint, R. Perioperative management and monitoring in anaesthesia. Lancet 2003, 362, 1839–1846. [Google Scholar] [CrossRef]
- Bendjelid, K.; Marx, G.; Kiefer, N.; Simon, T.P.; Geisen, M.; Hoeft, A.; Siegenthaler, N.; Hofer, C.K. Performance of a new pulse contour method for continuous cardiac output monitoring: Validation in critically ill patients. Br. J. Anaesth. 2013, 111, 573–579. [Google Scholar] [CrossRef] [Green Version]
- Biais, M.; Nouette-Gaulain, K.; Cottenceau, V.; Vallet, A.; Cochard, J.F.; Revel, P.; Sztark, F. Cardiac output measurement in patients undergoing liver transplantation: Pulmonary artery catheter versus uncalibrated arterial pressure waveform analysis. Anesth. Analg. 2008, 106, 1480–1486. [Google Scholar] [CrossRef]
- Moon, Y.J.; Moon, H.S.; Kim, D.S.; Kim, J.M.; Lee, J.K.; Shim, W.H.; Kim, S.H.; Hwang, G.S.; Choi, J.S. Deep learning-based stroke volume estimation outperforms conventional arterial contour method in patients with hemodynamic instability. J. Clin. Med. 2019, 8, 1419. [Google Scholar] [CrossRef] [Green Version]
- Kato, N.; Fadlullah, Z.M.; Mao, B.M.; Tang, F.; Akashi, O.; Inoue, T.; Mizutani, K. The deep learning vision for heterogeneous network traffic control: Proposal, challenges, and future perspective. IEEE Wirel. Commun. 2017, 24, 146–153. [Google Scholar] [CrossRef]
- Pourbabaee, B.; Roshtkhari, M.J.; Khorasani, K. Deep convolutional neural networks and learning ECG features for screening paroxysmal atrial fibrillation patients. IEEE Trans. Syst. Man Cybern. Syst. 2018, 48, 2095–2104. [Google Scholar] [CrossRef]
- Yao, Q.H.; Wang, R.X.; Fan, X.M.; Liu, J.K.; Li, Y. Multi-class Arrhythmia detection from 12-lead varied-length ECG using Attention-based Time-Incremental Convolutional Neural Network. Inform. Fusion 2020, 53, 174–182. [Google Scholar] [CrossRef]
- Acharya, U.R.; Fujita, H.; Lih, O.S.; Hagiwara, Y.; Tan, J.H.; Adam, M. Automated detection of arrhythmias using different intervals of tachycardia ECG segments with convolutional neural network. Inform. Sci. 2017, 405, 81–90. [Google Scholar] [CrossRef]
- Yildirim, O.; Plawiak, P.; Tan, R.S.; Acharya, U.R. Arrhythmia detection using deep convolutional neural network with long duration ECG signals. Comput. Biol. Med. 2018, 102, 411–420. [Google Scholar] [CrossRef]
- Qayyum, A.; Anwar, S.M.; Awais, M.; Majid, M. Medical image retrieval using deep convolutional neural network. Neurocomputing 2017, 266, 8–20. [Google Scholar] [CrossRef] [Green Version]
- Milletari, F.; Navab, N.; Ahmadi, S.-A. V-Net: Fully convolutional neural networks for volumetric medical image segmentation. arXiv 2016, arXiv:1606.04797. [Google Scholar]
- Lee, H.C.; Jung, C.W. Vital Recorder-a free research tool for automatic recording of high-resolution time-synchronised physiological data from multiple anaesthesia devices. Sci. Rep. 2018, 8, 1527. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Simonyan, K.; Zisserman, A. Very deep convolutional networks for large-scale image recognition. arXiv 2014, arXiv:1409.1556. [Google Scholar]
- Jyh-Cherng Gu, S.-L.Y. Removal of DC offset in current and voltage signals using a novel Fourier filter algorithm. IEEE Trans. Power Deliv. 2000, 15, 73–79. [Google Scholar] [CrossRef]
- Springenberg, J.T.; Dosovitskiy, A.; Brox, T.; Riedmiller, M. Striving for simplicity: The all convolutional net. arXiv 2014, arXiv:1412.6806. [Google Scholar]
- Bai, S.; Kolter, J.Z.; Koltun, V. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling. arXiv 2018, arXiv:1803.01271. [Google Scholar]
- Diederik, P.; Kingma, J.L.B. Adam: A Method for Stochastic Optimization. arXiv 2015, arXiv:1412.6980. [Google Scholar]
- Oh, S.L.; Ng, E.Y.K.; Tan, R.S.; Acharya, U.R. Automated diagnosis of arrhythmia using combination of CNN and LSTM techniques with variable length heart beats. Comput. Biol. Med. 2018, 102, 278–287. [Google Scholar] [CrossRef]
- Wu, T. CNN and RNN-based deep learning methods for digital signal demodulation. In Proceedings of the IVSP 2019: 2019 International Conference on Image, Video and Signal Processing, Shanghai, China, 25–28 February 2019; pp. 122–127. [Google Scholar]
- Li, Z.; Qin, Z.; Huang, K.; Yang, X.; Ye, S. Intrusion detection using convolutional neural networks for representation learning. In Neural Information Processing. ICONIP 2017, Lecture Notes in Computer Science; Liu, D., Xie, S., Li, Y., Zhao, D., El-Alfy, E.S., Eds.; Springer: Cham, Switzerland, 2017; pp. 858–866. [Google Scholar]
- Zhang, Z.; Lu, B.; Sheng, X.; Jin, N. Accuracy of stroke volume variation in predicting fluid responsiveness: A systematic review and meta-analysis. J. Anesth. 2011, 25, 904–916. [Google Scholar] [CrossRef] [PubMed]
Training Set (N = 210) | Validation Set (N = 217) | Test Set (N = 130) | Total Set (N = 557) | p-Value | |
---|---|---|---|---|---|
Demographics | |||||
Age (yrs.) | 58 (49–63) | 56 (45–63) | 56 (41–64) | 57 (47–63) | 0.17 |
Sex (male) | 145 (69.0%) | 154 (71.0%) | 82 (63.1%) | 381 (68.4%) | 0.30 |
Weight (kg) | 65 ± 12 | 67 ± 13 | 64 ± 12 | 66 ± 13 | 0.02 |
Height (cm) | 166 (160–171) | 167 (160–172) | 166 (160–170) | 166 (160–172) | 0.13 |
Body mass index (kg/m2) | 23.4 (21.1–26.2) | 24.0 (21.6–26.8) | 23.4 (21.6–25.4) | 23.6 (21.3–26.2) | 0.12 |
ASA classification a | <0.001 | ||||
1 | 12 (5.7) | 8 (3.7) | 23 (17.7) | 43 (7.7) | |
2 | 63 (30.0) | 68 (31.3) | 68 (52.3) | 199 (35.7) | |
3 | 111 (52.9) | 115 (53.0) | 38 (29.2) | 264 (47.4) | |
4 | 18 (8.6) | 25 (11.5) | 0 (0.0) | 43 (7.7) | |
5 | 6 (2.9) | 1 (0.5) | 1 (0.8) | 8 (1.4) | |
Underlying disease | |||||
Diabetes mellitus | 55 (26.2) | 56 (25.8) | 26 (20.0) | 137 (24.6) | 0.38 |
Hypertension | 63 (30.0) | 69 (31.8) | 32 (24.6) | 164 (29.4) | 0.36 |
Operation time, mins | 779 (399–870) | 755 (423–842) | 430 (320–740) | 733 (376–834) | <0.001 |
Emergency surgery | 25 (11.9) | 20 (9.2) | 8 (6.2) | 53 (9.5) | 0.21 |
Type of Operation | |||||
Transplant b | 150 (71.4) | 150 (69.1) | 53 (40.8) | 353 (63.4) | <0.001 |
Major open abdominal surgery | 52 (24.8) | 62 (28.6) | 71 (54.6) | 185 (33.2) | <0.001 |
Major laparoscopic abdominal surgery | 2 (1.0) | 4 (1.8) | 4 (3.1) | 10 (1.8) | 0.36 |
Minor abdominal surgery | 5 (2.4) | 0 (0.0) | 0 (0.0) | 5 (0.9) | 0.02 |
Others c | 1 (0.5) | 1 (0.5) | 2 (1.5) | 4 (0.7) | 0.45 |
Training Set (N = 210, 37.7%) | Validation Set (N = 217, 38.9%) | Test Set (N = 130, 23.4%) | Overall (N = 557) | p-Value | |
---|---|---|---|---|---|
Duration (min) | 120,679 | 131,474 | 31,598 | 283,752 | |
Blood pressure (mmHg) | |||||
Systolic | 110.7 ± 16.8 | 114.2 ± 17.4 | 113.0 ± 18.1 | 112.8 ± 17.4 | <0.001 |
Diastolic | 55.2 ± 9.6 | 57.5 ± 9.9 | 56.2 ± 9.9 | 56.5 ± 9.8 | <0.001 |
Heart rate (bpm) | 82.0 ± 15.2 | 81.3 ± 16.2 | 82.8 ± 14.1 | 81.8 ± 15.5 | <0.001 |
Stroke volume (mL/beat) | 87.3 ± 29.4 | 86.6 ± 26.4 | 80.0 ± 24.2 | 85.2 ± 27.5 | <0.001 |
Stroke volume index (mL/beat/m2) | 50.7 ± 16.2 | 50.1 ± 14.5 | 48.8 ± 15.2 | 50.2 ± 15.3 | <0.001 |
Systemic vascular resistance (dyne∙s/cm5) | 854.3 ± 354.5 | 880.7 ± 331.8 | 931.4 ± 381.0 | 877.6 ± 349.6 | <0.001 |
Cardiac output (L/min) | 7.0 ± 2.6 | 6.9 ± 2.3 | 6.5 ± 2.0 | 6.9 ± 2.4 | <0.001 |
Stroke volume variance (%) | 8.1 ± 4.9 | 8.1 ± 4.4 | 9.4 ± 5.2 | 8.3 ± 4.8 | <0.001 |
Model Type | Pearson Correlation, r | Mean Squared Error |
---|---|---|
Min-max Normalization + max pooling | 0.64 | 33.21 |
Min-max Normalization + average pooling | 0.66 | 22.86 |
Min-max Normalization + convolutional strides | 0.65 | 31.15 |
Removed DC offset + max pooling | 0.80 | 9.59 |
Removed DC offset + average pooling | 0.83 | 9.3 |
Removed DC offset + convolutional strides | 0.91 | 6.92 |
Data Type | Linear Regression Analysis | Bland-Altman Analysis | Mean Absolute Error | Mean Squared Error | Concordance Rate (%) | |
---|---|---|---|---|---|---|
Pearson Correlation | Bias | 95% Limits of Agreement | ||||
ABP signal | 0.91 | −1.00 | −4.47~2.48 | 1.55 | 4.74 | 90.14% |
Pre-processed ABP | 0.93 | −0.87 | −4.34~2.59 | 1.30 | 4.08 | 92.56% |
Frequency of ABP | 0.88 | −0.88 | −5.03~3.27 | 1.52 | 5.08 | 88.19% |
Slope of ABP | 0.93 | −1.00 | −4.62~2.61 | 1.38 | 4.59 | 92.99% |
Combined pre-processed and slope ABP | 0.94 | −0.93 | −4.02~2.17 | 1.24 | 3.18 | 92.86% |
Combined pre-processed, frequency and slope of ABP | 0.95 | −0.85 | −2.88~0.71 | 1.01 | 2.13 | 96.26% |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Kwon, H.-M.; Seo, W.-Y.; Kim, J.-M.; Shim, W.-H.; Kim, S.-H.; Hwang, G.-S. Estimation of Stroke Volume Variance from Arterial Blood Pressure: Using a 1-D Convolutional Neural Network. Sensors 2021, 21, 5130. https://doi.org/10.3390/s21155130
Kwon H-M, Seo W-Y, Kim J-M, Shim W-H, Kim S-H, Hwang G-S. Estimation of Stroke Volume Variance from Arterial Blood Pressure: Using a 1-D Convolutional Neural Network. Sensors. 2021; 21(15):5130. https://doi.org/10.3390/s21155130
Chicago/Turabian StyleKwon, Hye-Mee, Woo-Young Seo, Jae-Man Kim, Woo-Hyun Shim, Sung-Hoon Kim, and Gyu-Sam Hwang. 2021. "Estimation of Stroke Volume Variance from Arterial Blood Pressure: Using a 1-D Convolutional Neural Network" Sensors 21, no. 15: 5130. https://doi.org/10.3390/s21155130