Real Time Non-Invasive Hemodynamic Assessment of Ventricular Tachycardia

IEEE ACCESS(2020)

引用 0|浏览14
暂无评分
摘要
Hemodynamically unstable ventricular tachycardia (VT) is a critical cardiac arrhythmia associated with hemodynamic compromise that requires immediate cardioversion to prevent sudden cardiac death. Since unnecessary cardioverter defibrillators shocks damage the heart and increase the risk of mortality, the discrimination between unstable (i.e. requiring cardioversion) and stable (i.e. not requiring cardioversion) VT is of paramount importance. The aim of this study was to propose and assess non-invasive identification of hemodynamically unstable VT using photoplethysmography (PPG). Seventy-five (n = 75) episodes of VT were recorded in 14 patients undergoing invasive electrophysiological studies for VT catheter ablation. Invasive continuous arterial blood pressure (ABP), PPG and electrocardiogram (ECG) were simultaneously recorded. VTs were classified as unstable if during the first 10 seconds from onset, the mean ABP ((P-VT) over bar < 60<(P-VT)over bar>) was (P-VT) over bar (P-VT) over bar < 60 mmHg or if <(P-VT)over bar> dropped more than 30% with respect to a 10 seconds baseline (i.e. ratio R-ABP < 0.70 ). Five PPG morphological features were derived and compared to the heart rate from the ECG. PPG markers detected hemodynamically unstable VT with accuracy as high as 86% and were more accurate than the heart rate. The mean absolute slope was the best PPG parameter for classification of <(P-VT)over bar> < 60 <(P-VT)over bar> < 60 <(P-VT)over bar> < 60 mmHg (AUC = 0.85, Sensitivity = 72%, Specificity = 86%) and R-ABP < 0.70 R-ABP < 0.70 (AUC = 0.90, Sensitivity = 83%, Specificity = 89%) and it was automatically selected in the best two-variables logistic regression, for which AUC = 0.94. In conclusion, PPG analysis can accurately identify haemodynamically unstable VTs and has potential to enable optimization of VT therapy and reduce unnecessary and harmful cardioversion shocks.
更多
查看译文
关键词
Assistive technology,biomedical signal processing,cardiology
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要