Patient Outcome Prediction with Heart Rate Variability and Vital Signs

Abstract

The ability to predict patient outcomes is important for clinical triage, which is the process of assessing severity and assigning appropriate priority of treatment for large numbers of patients. In this study, we present an automatic prognosis system for patient outcome prediction with heart rate variability (HRV) and traditional vital signs. Support vector machine (SVM) and extreme learning machine (ELM) are employed as predictors, and SVM with linear kernel is reported to perform the best in general. In the experiments, the combination of HRV measures and vital signs is found to be more closely associated with patient outcome than either HRV or vital signs. Moreover, two new segment based methods are proposed to improve the predictive accuracy, where several sets of HRV measures are calculated from non-overlapped segments for each patient and final decision is made through the majority voting rule. The results reveal that the segment based methods are able to enhance the prediction performance significantly.

Prediction of cardiac arrest in critically ill patients presenting to the emergency department using a machine learning score incorporating heart rate variability compared with the modified early warning score

“…In our previous research, we proposed using a combination of age, HRV measures, and vital signs as a predictor of patient outcomes and demonstrated that the combined features present significant improvements to predictive accuracy, sensitivity, and specificity compared with using HRV alone [22,57]. As we can see from Table 3 not all of the vital signs were highly predictive when used in isolation.…”

“…We have investigated an extreme learning machine and a support vector machine with different activation/kernel functions as classifiers, and found that the linear support vector machine is able to provide the highest confidence in categorizing patients into two outcomes: death and survival. Furthermore, we have also presented a new segment-based decision-making strategy for outcome prediction [22]. …”

“…A ML-based prediction model - utilizing age, HRV parameters, and vital signs - was proposed to compute risk score on patient's hospital outcome [22]. This model was run on a MATLAB code (R2009a; The Mathworks).…”

Risk Scoring for Prediction of Acute Cardiac Complications from Imbalanced Clinical Data

“…ECG signals were sampled at a rate of 125 Hz and the processing of raw data to obtain the HRV parameters was done using the LABVIEW (Version 8.6, National Instruments, Austin, TX, USA) interface embedded with MATLAB (R2009a, The MathWorks, Natick, MA, USA) scripts. A detailed description of data acquisition and signal processing can be found in our previous works [15], [19], in which a threshold-plus-derivative method was used to detect the QRS complexes, and all ectopic and nonsinus beats were excluded in accordance with the guidelines outlined by the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology [20]. In this study, a total of 16 time domain and frequency domain HRV parameters were derived, which are elaborated in Table I. The 12-lead ECG was also measured at the ED with PageWriter TC Series Cardiograph (Philips, Amsterdam, Netherlands) and parameters were either automatically derived by the device or manually calculated by a doctor.…”

Detecting pathogen exposure during the non-symptomatic incubation period using physiological data

“…Past physiological signal-based early infection detection work has been heavily focused on systemic bacterial infection [30][31][32][33][34][35], and largely centered upon higher sampling rates of body core temperature [35,36], advanced analyses of stronglyconfounded signals such as heart rate variability [31][32][33] or social dynamics [37], or sensor data fusion from already symptomatic (febrile) individuals [38]. While great progress has been made in developing techniques for signal-based early warning of bacterial infections and other critical illnesses in a hospital setting [39][40][41][42], we are aware of only one prior effort to extend these techniques to viral infections or other communicable pathogens in non-clinical contexts using wearable sensor systems [43].…”