电子科技大学学报
電子科技大學學報
전자과기대학학보
Journal of University of Electronic Science and Technology of China
2015年
6期
951-955,960
,共6页
自动检测%心肺复苏%胸外按压%线性判别分析%胸阻抗
自動檢測%心肺複囌%胸外按壓%線性判彆分析%胸阻抗
자동검측%심폐복소%흉외안압%선성판별분석%흉조항
automatic detection%cardiopulmonary resuscitation (CPR)%chest compression (CC)%linear discriminant analysis (LDA)%transthoracic impedance (TTI)
为了自动识别胸阻抗(TTI)信号中的按压和通气波形，完成相关重要参数的计算，并结合先验知识和机器智能从而完成对心肺复苏质量的监测评估，提出了一种基于模式识别的胸阻抗信号自动检测算法。基于实验采集的猪的电诱导心脏骤停模型TTI信号，设计结合小波和形态学的除噪算法对信号进行预处理，再由多分辨率窗口搜索法完成潜在按压和通气波形的定位，最后采用线性判别分析法对定位的按压和通气波形进行分类识别。实验结果表明，该算法对TTI信号中按压波形和波形分析识别的正确率和敏感度可达到98.237%、94.947%和99.651%、97.282%，稳定性好，且运行时间(0.485±0.07 s)满足实时性要求。
為瞭自動識彆胸阻抗(TTI)信號中的按壓和通氣波形，完成相關重要參數的計算，併結閤先驗知識和機器智能從而完成對心肺複囌質量的鑑測評估，提齣瞭一種基于模式識彆的胸阻抗信號自動檢測算法。基于實驗採集的豬的電誘導心髒驟停模型TTI信號，設計結閤小波和形態學的除譟算法對信號進行預處理，再由多分辨率窗口搜索法完成潛在按壓和通氣波形的定位，最後採用線性判彆分析法對定位的按壓和通氣波形進行分類識彆。實驗結果錶明，該算法對TTI信號中按壓波形和波形分析識彆的正確率和敏感度可達到98.237%、94.947%和99.651%、97.282%，穩定性好，且運行時間(0.485±0.07 s)滿足實時性要求。
위료자동식별흉조항(TTI)신호중적안압화통기파형，완성상관중요삼수적계산，병결합선험지식화궤기지능종이완성대심폐복소질량적감측평고，제출료일충기우모식식별적흉조항신호자동검측산법。기우실험채집적저적전유도심장취정모형TTI신호，설계결합소파화형태학적제조산법대신호진행예처리，재유다분변솔창구수색법완성잠재안압화통기파형적정위，최후채용선성판별분석법대정위적안압화통기파형진행분류식별。실험결과표명，해산법대TTI신호중안압파형화파형분석식별적정학솔화민감도가체도98.237%、94.947%화99.651%、97.282%，은정성호，차운행시간(0.485±0.07 s)만족실시성요구。
In order to recognize the compression and ventilation waveforms, obtain the important parameters, and evaluate the CPR quality by combining with prior knowledge, this paper proposes an automatic detection algorithm for transthoracic impedance (TTI) signal based on pattern recognition. The TTI signals that come from pig model based on electrically induced cardiac arrest are reprocessed by denoising algorithm based on wavelet and morphology firstly. Then the potential compression and ventilation waveforms are located by using the searching algorithm of multiresolution window. Finally, the linear discriminant analysis algorithm is used to classify and recognize the located compression and ventilation waveforms. The results show that both the recognition accuracy and sensitivity of the compression and ventilation waveforms are 98.237%, 94.947% and 99.651%, 97.282%, and the running time (0.485±0.07s) satisfies the requirement of clinical applications.
