基於手腕光體積描記訊號於通用化及個人化血壓估測模型之應用

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：47

、訪客IP：18.223.159.141

姓名

楊家毓(Jia-Yu Yang) 查詢紙本館藏

畢業系所

電機工程學系

論文名稱

基於手腕光體積描記訊號於通用化及個人化血壓估測模型之應用
(Application of Generalized and Personalized Blood Pressure Estimation Models based on Wrist Photoplethymography Signals)

相關論文

★ 具輸出級誤差消除機制之三位階三角積分D類放大器設計	★ 應用於無線感測網路之多模式低複雜度收發機設計
★ 用於數位D類放大器的高效能三角積分調變器設計	★ 交換電容式三角積分D類放大器電路設計
★ 適用於平行處理及排程技術的無衝突定址法演算法之快速傅立葉轉換處理器設計	★ 適用於IEEE 802.11n之4×4多輸入多輸出偵測器設計
★ 應用於無線通訊系統之同質性可組態記憶體式快速傅立葉處理器	★ 3GPP LTE正交分頻多工存取下行傳輸之接收端細胞搜尋與同步的設計與實現
★ 應用於3GPP-LTE下行多天線接收系統高速行駛下之通道追蹤與等化	★ 適用於正交分頻多工系統多輸入多輸出訊號偵測之高吞吐量QR分解設計
★ 應用於室內極高速傳輸無線傳輸系統之設計與評估	★ 適用於3GPP LTE-A之渦輪解碼器硬體設計與實作
★ 下世代數位家庭之千兆級無線通訊系統	★ 協作式通訊於超寬頻通訊系統之設計
★ 適用於3GPP-LTE系統高行車速率基頻接收機之設計	★ 多使用者多輸入輸出前編碼演算法及關鍵組件設計

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 (2028-8-31以後開放)

摘要(中)

心血管疾病對於人類健康造成的威脅不容小覷，而血壓作為心血管疾病的重要指標，臨床上通常以血壓計配合壓脈帶進行量測，然而該方式除了有諸多不便之處，且容易受到白袍效應而無法得到客觀的數值外，亦無法連續地長時間追蹤數值。近年來受惠於穿戴式裝置的發展，許多研究紛紛投入無脈壓帶的血壓估測方法，卻在準確度、應用環境及實用性等層面面臨挑戰。本研究利用穿戴式手錶可量測的光體積描記訊號來估測血壓數值，並依照校正與否及應用範圍分為通用化模型及個人化模型。從裝置量測的訊號處理階段開始，先對波形進行前處理並去除基線飄移及市電訊號的干擾，接著設定預篩選條件過濾品質不良的訊號，再透過特徵提取與波形拆解來解析訊號特性與標記特徵值，最後利用訊號品質檢測確保提取的特徵具有代表性。在資料處理階段，對於無須校正的通用化模型，我們將資料分成訓練、驗證及測試資料，並對訓練資料進行資料擴充，避免模型因不均勻的資料分布導致估測偏差，而測試資料則依照美國醫療儀器促進協會(AAMI)的標準，選擇85人255筆且符合血壓區間規範之資料集，確保效能估測的客觀性。我們嘗試了三種機器學型模型，包含神經網路、卷積神經網路及殘差網路，經由驗證的估測標準差，在收縮壓上達到11.35(mmHg)、舒張壓達到8.75(mmHg)的準確度，在相同條件且符合AAMI驗證標準下優於其他文獻。對於有校正之個人化模型，我們利用遷移式學習的概念來輔助少樣本學習的應用條件，並利用模型無關之元學習演算法建立個人化模型，在受試者可自由活動及運動的前提估測其日間及夜間血壓，收縮壓之標準差達到 7.78(mmHg)，舒張壓之標準差達到6.99(mmHg)，夜間血壓下降之絕對誤差達4.45(mmHg)，我們在量測訊號種類較少、裝置較簡便的前提優於文獻上其他也是自由活動情境下之個人化模型。

摘要(英)

The threat of cardiovascular disease to human health cannot be underestimated. As an important indicator of cardiovascular disease, blood pressure is usually measured clinically with a sphygmomanometer and a cuff. In addition to being unable to obtain objective value due to white-coat effect, it is also impossible to continuously track the value for a long time in daily life. In recent years, benefiting from the development of wearable device, many studies have invested in cuffless blood pressure estimation methods, but they face challenges in terms of accuracy, application environment and practicability. This study uses the photoplethysmography from the smart watch to estimate the blood pressure. A general model and a personal model are constructed respectively according to whether it is calibrated or not and the application. Starting from the signal processing stage, the waveform is pre-processed to remove baseline wandering and the interference, then the pre-screening conditions are set to remove the signal with poor quality. The charactersitics of PPG morphology are acquired through feature extraction and waveform decomposition, and finally signal quality assessment is used to ensure that the extracted features are representative. In the data processing stage, for calibration-free general model, we divide the data into training, validation and testing sets, and augment the training data to avoid estimation bias. The testing data is in accordance with with the standard of the Association for the Advancement of Medical Instrumentation (AAMI), selecting 85 people with 255 data that meet the blood pressure interval specifications of required distribution. We tried three machine learning models, including neural network, convolutional neural network and residual network, the estimated standard deviation reaches 11.35 (mmHg) and 8.75 (mmHg) for systolic and diastolic blood pressure. The accuracy is better than other literature with wrist PPG signals under the same conditions and in compliance with AAMI verification standards. For personal model with calibration, we use the concept of transfer learning to assist the application conditions of few-shot learning, and use the model-agnostic meta-learning algorithm to establish initial weights of the personal model, the standard deviation reached 7.78 (mmHg) and 6.99 (mmHg) for systolic and diastolic blood pressure respectively, and the mean absolute error of nocturnal blood pressure dip reached 4.45 (mmHg). The results outperform the 24-hour free-living personal model in the literature.

關鍵字(中)

★ 光體積變化描記訊號
★ 高血壓
★ 無關模型之元學習
★ 收縮壓
★ 舒張壓
★ 三軸加速度計訊號

關鍵字(英)

★ Photoplethysmography Signal
★ Hypertension
★ Model-Agnostic Meta-Learning
★ Systolic Blood Pressure
★ Diastolic Blood Pressure
★ Triaxial Accelerator Signal

論文目次

摘要---i
Abstract---ii
目錄---iii
表目錄---vi
圖目錄---viii
第一章緒論---1
1.1 研究動機---1
1.2 研究方法---1
1.3 論文組織---2
第二章穿戴式裝置訊號---3
2.1 收縮壓與舒張壓(Systolic and Diastolic Blood Pressure)---3
2.2 光體積變化描記訊號(Photoplethysmography, PPG)---4
2.3 三軸加速度計訊號(Triaxial Accelerator Signal)---5
第三章光體積描記訊號處理---6
3.1 流程圖---6
3.2 二十四小時訊號前處理---7
3.2.1 範圍提取(Select Range)---7
3.2.2 檢測氣囊充氣(Detect Cuff Inflation)---8
3.2.3 確認氣囊充氣時間(Check Cuff Inflation Time)---9
3.2.4 計算活動指數(Activity Index, AI)---11
3.2.5 活動計數(Activity Counts)計算與驗證---12
3.2.6 預處理及升取樣(Preprocessing and Up-sampling)---14
3.3 靜態訊號前處理---16
3.4 預篩選(Pre-screening, PreSQI)---16
3.4.1 飽和訊號檢測(Saturated Signal Detection)---17
3.4.2 異常幅距檢測(Abnormal Range Detection)---18
3.5 特徵提取(Feature Extraction)---19
3.5.1 光體積變化描記訊號特徵提取---19
3.6 訊號品質檢測(Signal Quality Index, SQI)---20
3.7 加權式波型拆解(Weighted Pulse Decomposition)---21
3.7.1 波形拆解---22
3.7.2 波型拆解品質檢測(Waveform Decomposition SQI)---23
第四章資料處理---24
4.1 流程圖---24
4.2 資料可靠度檢查(Signal Reliability Check)---25
4.3 計算K值與多尺度熵(K Value and Multiscale Entropy)---25
4.4 波形提取(Waveform Extraction)---26
4.5 特徵可靠度檢查(Feature Reliability Check)---29
4.6 特徵擴充(Feature Combination)---29
4.7 取中位數(Median)---31
第五章通用化模型血壓估測及結果---32
5.1 資料集分割(Data Set Segmentation)---32
5.1.1 無波形模型資料集分割---32
5.1.2 含波形模型資料集分割---36
5.2 資料擴增(Data Augmentation)---39
5.3 通用化模型(General)演算法---41
5.3.1 神經網路(Neural Network, NN)---41
5.3.2 卷積神經網路(Convolutional Neural Network, CNN)---44
5.3.3 殘差網路(Residual Network, ResNet)---48
5.4 通用化模型結果---52
5.4.1 神經網路結果---53
5.4.2 卷積神經網路結果---57
5.4.3 殘差網路結果---60
5.5 與其他論文比較---63
5.6 總結---67
第六章個人化模型血壓估測及結果---68
6.1 個人化模型演算法---68
6.1.1 遷移式學習(Transfer Learning)---68
6.1.2 少樣本學習(Few-Shot Learning)及元學習(Meta Learning)---69
6.1.3 無關模型之元學習演算法(Model-Agnostic Meta-Learning)[35]---71
6.2 資料集分割---75
6.2.1 訓練任務---75
6.2.2 測試任務---79
6.3 個人化模型結果---87
6.3.1 個人化模型---87
6.3.2 方法一(method1)之估測結果---90
6.3.3 方法二(method2)之估測結果---93
6.3.4 方法三(method3)之估測結果---96
6.4 比較與總結---99
第七章總結與未來展望---103
參考資料---104

參考文獻

[1] Smith SC Jr, Collins A, Ferrari R et al, World Heart Federation; American Heart Association; American College of Cardiology Foundation; European Heart Network; European Society of Cardiology. Our time: a call to save preventable death from cardiovascular disease (heart disease and stroke). J Am Coll Cardiol. 2012 Dec 4;60(22):2343-8. doi: 10.1016/j.jacc.2012.08.962. Epub 2012 Sep 18. PMID: 22995536.
[2] Lawes CM, Bennett DA, Parag V et al; Asia Pacific Cohort Studies Collaboration. Blood pressure indices and cardiovascular disease in the Asia Pacific region: a pooled analysis. Hypertension. 2003 Jul;42(1):69-75. doi: 10.1161/01.HYP.0000075083.04415.4B. Epub 2003 May 19. PMID: 12756223.
[3] O′Brien E, Parati G, Stergiou G et al; European Society of Hypertension Working Group on Blood Pressure Monitoring. European Society of Hypertension position paper on ambulatory blood pressure monitoring. J Hypertens. 2013 Sep;31(9):1731-68. doi: 10.1097/HJH.0b013e328363e964. Erratum in: J Hypertens. 2013 Dec;31(12):2467. PMID: 24029863.
[4] Walker HK, Hall WD, Hurst JW, editors. Clinical Methods: The History, Physical, and Laboratory Examinations. 3rd ed. Boston: Butterworths; 1990. PMID: 21250045.
[5] K. Song, K. -y. Chung and J. -H. Chang, "Cuffless Deep Learning-Based Blood Pressure Estimation for Smart Wristwatches," in IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 7, pp. 4292-4302, July 2020, doi: 10.1109/TIM.2019.2947103.
[6] Pribadi, Eka Fitrah & Pandey, Rajeev Kumar & Chao, Paul. (2020). Optimizing a novel PPG sensor patch via optical simulations towards accurate heart rates. Microsystem Technologies. 26. 10.1007/s00542-020-04895-6.
[7] https://www.philips.com.au/healthcare/product/HCNOCTN446/actiwatch-spectrum-pro-get-the-actiwatch-advantage
[8] Bai, Jiawei & Di, Chongzhi & Xiao, Luo & Evenson, Kelly & Lacroix, Andrea & Crainiceanu, Ciprian & Buchner, David. (2016). An Activity Index for Raw Accelerometry Data and Its Comparison with Other Activity Metrics. PLOS ONE. 11. 10.1371/journal.pone.0160644.
[9] ActiGraph. What are counts? [Internet]. 2011 [cited 1 Oct 2015].https://actigraphcorp.my.site.com/support/s/article/What-are-counts
[10] 郭嘉威(109)。，”基於手指光電容積描記法與心電訊號之血壓估測”，國立中央大學電機工程學系，桃園市。
[11] Tsai, P.-Y. et al, “Coherence between decomposed components of wrist and finger PPG signals by imputing missing features and resolving ambiguous features.”, Sensors 21, 4315, 2021.
[12] Couceiro R, Carvalho P, Paiva RP, Henriques J, Quintal I, Antunes M, Muehlsteff J, Eickholt C, Brinkmeyer C, Kelm M, Meyer C. Assessment of cardiovascular function from multi-Gaussian fitting of a finger photoplethysmogram. Physiol Meas. 2015 Sep;36(9):1801-25. doi: 10.1088/0967-3334/36/9/1801. Epub 2015 Aug 3. PMID: 26235798.
[13] 黃秋樺(108)。，”基於光電容積描記法之手腕脈波傳導速度估測”，國立中央大學電機工程學系，桃園市。.
[14] Q. Wu et al., "An Ambulatory Blood Pressure Monitoring System Based on the Uncalibrated Steps of the Wrist," 2019 12th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI), Suzhou, China, 2019, pp. 1-6, doi: 10.1109/CISP-BMEI48845.2019.8966011.
[15] Costa M, Goldberger AL, Peng CK. Multiscale entropy analysis of biological signals. Phys Rev E Stat Nonlin Soft Matter Phys. 2005 Feb;71(2 Pt 1):021906. doi: 10.1103/PhysRevE.71.021906. Epub 2005 Feb 18. PMID: 15783351.
[16] Stergiou GS, Alpert B, Mieke S et al, “A universal standard for the validation of blood pressure measuring devices: Association for the Advancement of Medical Instrumentation/European Society of Hypertension/International Organization for Standardization (AAMI/ESH/ISO) Collaboration Statement,” J Hypertens. 2018 Mar;36(3):472-478. doi: 10.1097/HJH.0000000000001634. PMID: 29384983; PMCID: PMC5796427.
[17] Glorot, Xavier & Bordes, Antoine & Bengio, Y.. (2010). Deep Sparse Rectifier Neural Networks. Journal of Machine Learning Research. 15.
[18] A. A. M. Al-Saffar, H. Tao and M. A. Talab, "Review of deep convolution neural network in image classification," 2017 International Conference on Radar, Antenna, Microwave, Electronics, and Telecommunications (ICRAMET), Jakarta, Indonesia, 2017, pp. 26-31, doi: 10.1109/ICRAMET.2017.8253139.
[19] R. Xin, J. Zhang and Y. Shao, "Complex network classification with convolutional neural network," in Tsinghua Science and Technology, vol. 25, no. 4, pp. 447-457, Aug. 2020, doi: 10.26599/TST.2019.9010055.
[20] 李御銓(110)。，”基於階層迴歸方法運用光體積描記訊號與心電圖估測血壓與脈波傳導速度”，國立中央大學電機工程學系，桃園市。
[21] K. He, X. Zhang, S. Ren and J. Sun, "Deep Residual Learning for Image Recognition," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 2016 pp. 770-778. doi: 10.1109/CVPR.2016.90
[22] Kingma, Diederik & Ba, Jimmy. (2014). Adam: A Method for Stochastic Optimization. International Conference on Learning Representations.
[23] Tieleman, T. and Hinton, G. (2012) Lecture 6.5-rmsprop: Divide the Gradient by a Running Average of Its Recent Magnitude. COURSERA: Neural Networks for Machine Learning, 4, 26-31.
[24] Duchi, John & Hazan, Elad. (2011). Adaptive Subgradient Methods for Online Learning and Stochastic Optimization. Journal of Machine Learning Research. 12. 2121-2159.
[25] S. Baek, J. Jang and S. Yoon, "End-to-End Blood Pressure Prediction via Fully Convolutional Networks," in IEEE Access, vol. 7, pp. 185458-185468, 2019, doi: 10.1109/ACCESS.2019.2960844.
[26] Z. Li et al., "A Novel Method for Calibration-Based Cuff-Less Blood Pressure Estimation," 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Berlin, Germany, 2019, pp. 4266-4269, doi: 10.1109/EMBC.2019.8857373.
[27] Chowdhury MH, Shuzan MNI, Chowdhury MEH, Mahbub ZB, Uddin MM, Khandakar A, Reaz MBI. Estimating Blood Pressure from the Photoplethysmogram Signal and Demographic Features Using Machine Learning Techniques. Sensors (Basel). 2020 Jun 1;20(11):3127. doi: 10.3390/s20113127. PMID: 32492902; PMCID: PMC7309072.
[28] M. Panwar, A. Gautam, D. Biswas and A. Acharyya, "PP-Net: A Deep Learning Framework for PPG-Based Blood Pressure and Heart Rate Estimation," in IEEE Sensors Journal, vol. 20, no. 17, pp. 10000-10011, 1 Sept.1, 2020, doi: 10.1109/JSEN.2020.2990864.
[29] B. Paliakaitė et al., "Blood Pressure Estimation Based on Photoplethysmography: Finger Versus Wrist," 2021 Computing in Cardiology (CinC), Brno, Czech Republic, 2021, pp. 1-4, doi: 10.23919/CinC53138.2021.9662716.
[30] Shen, Chia-Hao et al. “Language Transfer of Audio Word2Vec: Learning Audio Segment Representations Without Target Language Data.” 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (2017): 2231-2235.
[31] J. -T. Huang, J. Li, D. Yu, L. Deng and Y. Gong, "Cross-language knowledge transfer using multilingual deep neural network with shared hidden layers," 2013 IEEE International Conference on Acoustics, Speech and Signal Processing, Vancouver, BC, Canada, 2013, pp. 7304-7308, doi: 10.1109/ICASSP.2013.6639081.
[32] O. Simeone, S. Park and J. Kang, "From Learning to Meta-Learning: Reduced Training Overhead and Complexity for Communication Systems," 2020 2nd 6G Wireless Summit (6G SUMMIT), Levi, Finland, 2020, pp. 1-5, doi: 10.1109/6GSUMMIT49458.2020.9083856.
[33] Hung-yi Lee, Shang-Wen Li, and Thang Vu. 2022. Meta Learning for Natural Language Processing: A Survey. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 666–684, Seattle, United States. Association for Computational Linguistics.
[34] https://www.youtube.com/watch?v=EkAqYbpCYAc
[35] Finn, Chelsea and Abbeel, Pieter and Levine, Sergey, “Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks”, arXiv, 2017, doi: 0.48550/ARXIV.1703.03400
[36] Paschos GK, Baggs JE, Hogenesch JB, FitzGerald GA. The role of clock genes in pharmacology. Annu Rev Pharmacol Toxicol. 2010;50:187-214. doi: 10.1146/annurev.pharmtox.010909.105621. PMID: 20055702.
[37] Radha M, de Groot K, Rajani N, Wong CCP, Kobold N, Vos V, Fonseca P, Mastellos N, Wark PA, Velthoven N, Haakma R, Aarts RM. Estimating blood pressure trends and the nocturnal dip from photoplethysmography. Physiol Meas. 2019 Feb 26;40(2):025006. doi: 10.1088/1361-6579/ab030e. PMID: 30699397.
[38] Y. -L. Zheng, B. P. Yan, Y. -T. Zhang and C. C. Y. Poon, "An Armband Wearable Device for Overnight and Cuff-Less Blood Pressure Measurement," in IEEE Transactions on Biomedical Engineering, vol. 61, no. 7, pp. 2179-2186, July 2014, doi: 10.1109/TBME.2014.2318779.
[39] Sun S, Bezemer R, Long X, Muehlsteff J, Aarts RM. Systolic blood pressure estimation using PPG and ECG during physical exercise. Physiol Meas. 2016 Dec;37(12):2154-2169. doi: 10.1088/0967-3334/37/12/2154. Epub 2016 Nov 14. PMID: 27841157.
[40] Ruiz-Rodríguez, Juan & Ruiz-Sanmartín, Adolf & Ribas, Vicent & Caballero, Jesus & García-Roche, Alejandra & Riera, Jordi & Nuvials, Xavier & de Nadal, Miriam & De Sola-Morales, Oriol & Serra, Joaquim & Rello, Jordi. (2013). Innovative continuous non-invasive cuffless blood pressure monitoring based on photoplethysmography technology. Intensive care medicine. 39. 10.1007/s00134-013-2964-2.

指導教授

蔡佩芸(Pei-Yun Tsai)

審核日期

2023-8-9

推文