Patient Outcome Prediction with Heart Rate Variability and Vital Signs

Objective: Vagus nerve stimulation (VNS) is an adjunctive and well-established treatment for patients with drug-resistant epilepsy (DRE). However, it is still difficult to identify patients who may benefit from VNS surgery. Our study aims to propose a VNS outcome prediction model based on machine learning with multidimensional preoperative heart rate variability (HRV) indices.Methods: The preoperative electrocardiography (ECG) of 59 patients with DRE and of 50 healthy controls were analyzed. Responders were defined as having at least 50% average monthly seizure frequency reduction at 1-year follow-up. Time domain, frequency domain, and non-linear indices of HRV were compared between 30 responders and 29 non-responders in awake and sleep states, respectively. For feature selection, univariate filter and recursive feature elimination (RFE) algorithms were performed to assess the importance of different HRV indices to VNS outcome prediction and improve the classification performance. Random forest (RF) was used to train the classifier, and leave-one-out (LOO) cross-validation was performed to evaluate the prediction model.Results: Among 52 HRV indices, 49 showed significant differences between DRE patients and healthy controls. In sleep state, 35 HRV indices of responders were significantly higher than those of non-responders, while 16 of them showed the same differences in awake state. Low-frequency power (LF) ranked first in the importance ranking results by univariate filter and RFE methods, respectively. With HRV indices in sleep state, our model achieved 74.6% accuracy, 80% precision, 70.6% recall, and 75% F1 for VNS outcome prediction, which was better than the optimal performance in awake state (65.3% accuracy, 66.4% precision, 70.5% recall, and 68.4% F1).Significance: With the ECG during sleep state and machine learning techniques, the statistical model based on preoperative HRV could achieve a better performance of VNS outcome prediction and, therefore, help patients who are not suitable for VNS to avoid the high cost of surgery and possible risks of long-term stimulation.

Download Full-text

Integrating heart rate variability, vital signs, electrocardiogram, and troponin to triage chest pain patients in the ED

The American Journal of Emergency Medicine ◽

10.1016/j.ajem.2017.07.054 ◽

2018 ◽

Vol 36 (2) ◽

pp. 185-192 ◽

Cited By ~ 2

Author(s):

Jeffrey Tadashi Sakamoto ◽

Nan Liu ◽

Zhi Xiong Koh ◽

Dagang Guo ◽

Micah Liam Arthur Heldeweg ◽

...

Keyword(s):

Heart Rate ◽

Heart Rate Variability ◽

Chest Pain ◽

Vital Signs ◽

Pain Patients

Download Full-text

Heart Rate Variability Analysis Is More Sensitive At Identifying Neonatal Sepsis Than Conventional Vital Signs

Journal of Surgical Research ◽

10.1016/j.jss.2013.11.244 ◽

2014 ◽

Vol 186 (2) ◽

pp. 507

Author(s):

A.A. Mrazek ◽

M.T. Shabana ◽

G.L. Radhakrishnan ◽

G.C. Kramer ◽

R.S. Radhakrishnan

Keyword(s):

Heart Rate ◽

Heart Rate Variability ◽

Neonatal Sepsis ◽

Vital Signs ◽

Heart Rate Variability Analysis ◽

Variability Analysis

Download Full-text

Early prediction of serious infections in febrile infants incorporating heart rate variability in an emergency department: a pilot study

Emergency Medicine Journal ◽

10.1136/emermed-2020-210675 ◽

2021 ◽

pp. emermed-2020-210675

Author(s):

Shu-Ling Chong ◽

Gene Yong-Kwang Ong ◽

John Carson Allen ◽

Jan Hau Lee ◽

Rupini Piragasam ◽

...

Keyword(s):

Heart Rate ◽

Heart Rate Variability ◽

Vital Signs ◽

Model Performance ◽

Diagnostic Challenge ◽

Poincaré Plot ◽

Time Frequency ◽

Young Infants ◽

Serious Infections ◽

Non Linear

BackgroundEarly differentiation of febrile young infants with from those without serious infections (SIs) remains a diagnostic challenge. We sought to (1) compare vital signs and heart rate variability (HRV) parameters between febrile infants with versus without SIs, (2) assess the performance of HRV and vital signs with reference to current triage tools and (3) compare HRV and vital signs to HRV, vital signs and blood biomarkers, when predicting for the presence of SIs.MethodsUsing a prospective observational design, we recruited patients <3 months old presenting to a tertiary paediatric ED in Singapore from December 2018 through November 2019. We obtained patient demographic characteristics, triage assessment (including the Severity Index Score (SIS)), HRV parameters (time, frequency and non-linear domains) and laboratory results. We performed multivariable logistic regression analyses to predict the presence of an SI, using area under the curve (AUC) with the corresponding 95% CI to assess predictive capability.ResultsAmong 203 infants with a mean age of 38.4 days (SD 27.6), 67 infants (33.0%) had an SI. There were significant differences in the time, frequency and non-linear domains of HRV parameters between infants with versus without SIs. In predicting SIs, gender, temperature and the HRV non-linear parameter Poincaré plot SD2 (AUC 0.78, 95% CI 0.71 to 0.84) performed better than SIS alone (AUC 0.61, 95% CI 0.53 to 0.68). Model performance improved with the addition of absolute neutrophil count and C reactive protein (AUC 0.82, 95% CI 0.76 to 0.89).ConclusionAn exploratory prediction model incorporating HRV and biomarkers improved prediction of SIs. Further research is needed to assess if HRV can identify which young febrile infants have an SI at ED triage.Trial registration numberNCT04103151.

Download Full-text

Abstract 339: A Novel Continuous Heart Rate Variability Based Analytic to Predict Clinical Hemodynamic Instability Requiring Intervention

Circulation ◽

10.1161/circ.138.suppl_2.339 ◽

2018 ◽

Vol 138 (Suppl_2) ◽

Author(s):

Ashwin Belle ◽

Kevin R Ward ◽

Bryce Benson ◽

Mark Salamango ◽

Fadi Islim

Keyword(s):

Machine Learning ◽

Heart Rate ◽

Heart Rate Variability ◽

Lead Time ◽

Vital Signs ◽

Hemodynamic Instability ◽

Lead Times ◽

Compensatory Mechanisms ◽

Predictive Values ◽

Life Saving

Background: Sudden in-hospital hemodynamic instability (HI) due to cardiovascular and/or cardiorespiratory distress is a common occurrence. Causes can include hemorrhage, sepsis, pneumonia, heart failure, and others. Due to the body’s compensatory mechanisms, heart and respiratory rate, and blood pressure can be late indicators of HI. When detected late or left unrecognized HI can lead to complications and even death. Heart rate variability (HRV) has been demonstrated to reflect the status of the autonomic nervous system with changes in HRV linked to HI. However, to date, HRV has not demonstrated sufficient accuracies in adults to warrant widespread adoption. We have developed a novel nonlinear single lead ECG HRV analytic based on signal processing and machine learning features specific to HI physiology. Objective: Validate the capability, accuracy, and lead times of the HRV analytic to predict HI in hospitalized patients who were subjects of Rapid Response Team (RRT) activations. Methods: We retrospectively analyzed 4483 hours of ECG data from 22 RRT patient cases (16 male, 6 female) using the previously developed HRV analytic. A multi-clinician review adjudicated the occurrence of HI and need for intervention. The prediction lead time prior to the RRT call was calculated for these cases. Results: Of the 22 RRT cases, 13 calls were due to HI requiring a range of life-saving interventions, and 9 RRT cases for reasons not associated with HI and requiring no life-saving interventions. The analytic correctly distinguished between HI vs. non-HI RRT calls with 100% accuracy thus displaying 100% positive and negative predictive values. In the HI cases, the analytic detected HI with a median lead time of 7.7 hours prior to the RRT call with a range of 14 minutes to 38.4 hours. Conclusion: In this RRT cohort, a novel HRV analytic developed through machine learning based on HI physiology demonstrated its potential to forecast HI well before vital signs and other factors led to RRT activation resulting in life-saving interventions. These substantial lead times, as well as the ability to distinguish HI from non-HI RRT calls, may lead to the ability to reduce HI and improve outcomes while reducing unnecessary RRT activations.

Download Full-text

Utility of Vital Signs, Heart Rate Variability and Complexity, and Machine Learning for Identifying the Need for Lifesaving Interventions in Trauma Patients

Shock ◽

10.1097/shk.0000000000000186 ◽

2014 ◽

Vol 42 (2) ◽

pp. 108-114 ◽

Cited By ~ 23

Author(s):

Nehemiah T. Liu ◽

John B. Holcomb ◽

Charles E. Wade ◽

Mark I. Darrah ◽

Jose Salinas

Keyword(s):

Machine Learning ◽

Heart Rate ◽

Heart Rate Variability ◽

Vital Signs ◽

Trauma Patients

Download Full-text

Heart Rate Variability Predicts Trauma Patient Outcome as Early as 12 h: Implications for Military and Civilian Triage

Journal of Surgical Research ◽

10.1016/j.jss.2005.04.024 ◽

2005 ◽

Vol 129 (1) ◽

pp. 122-128 ◽

Cited By ~ 64

Author(s):

Patrick R. Norris ◽

John A. Morris ◽

Asli Ozdas ◽

Eric L. Grogan ◽

Anna E. Williams

Keyword(s):

Patient Outcome ◽

Heart Rate ◽

Heart Rate Variability ◽

Trauma Patient ◽

Trauma Patient Outcome

Download Full-text

Real-Time Webcam Heart-Rate and Variability Estimation with Clean Ground Truth for Evaluation

Applied Sciences ◽

10.3390/app10238630 ◽

2020 ◽

Vol 10 (23) ◽

pp. 8630

Author(s):

Amogh Gudi ◽

Marian Bittner ◽

Jan van Gemert

Keyword(s):

Heart Rate ◽

Heart Rate Variability ◽

Real Time ◽

Vital Signs ◽

Ground Truth ◽

Heart Beat ◽

Psychological Conditions ◽

Measure Heart Rate ◽

High Degree ◽

Motion Suppression

Remote photo-plethysmography (rPPG) uses a camera to estimate a person’s heart rate (HR). Similar to how heart rate can provide useful information about a person’s vital signs, insights about the underlying physio/psychological conditions can be obtained from heart rate variability (HRV). HRV is a measure of the fine fluctuations in the intervals between heart beats. However, this measure requires temporally locating heart beats with a high degree of precision. We introduce a refined and efficient real-time rPPG pipeline with novel filtering and motion suppression that not only estimates heart rates, but also extracts the pulse waveform to time heart beats and measure heart rate variability. This unsupervised method requires no rPPG specific training and is able to operate in real-time. We also introduce a new multi-modal video dataset, VicarPPG 2, specifically designed to evaluate rPPG algorithms on HR and HRV estimation. We validate and study our method under various conditions on a comprehensive range of public and self-recorded datasets, showing state-of-the-art results and providing useful insights into some unique aspects. Lastly, we make available CleanerPPG, a collection of human-verified ground truth peak/heart-beat annotations for existing rPPG datasets. These verified annotations should make future evaluations and benchmarking of rPPG algorithms more accurate, standardized and fair.

Download Full-text