Using support vector machines on photoplethysmographic signals to discriminate between hypovolemia and euvolemia

Identifying trauma patients at risk of imminent hemorrhagic shock is a challenging task in intraoperative and battlefield settings given the variability of traditional vital signs, such as heart rate and blood pressure, and their inability to detect blood loss at an early stage. To this end, we acquired N = 58 photoplethysmographic (PPG) recordings from both trauma patients with suspected hemorrhage admitted to the hospital, and healthy volunteers subjected to blood withdrawal of 0.9 L. We propose four features to characterize each recording: goodness of fit (r2), the slope of the trend line, percentage change, and the absolute change between amplitude estimates in the heart rate frequency range at the first and last time points. Also, we propose a machine learning algorithm to distinguish between blood loss and no blood loss. The optimal overall accuracy of discriminating between hypovolemia and euvolemia was 88.38%, while sensitivity and specificity were 88.86% and 87.90%, respectively. In addition, the proposed features and algorithm performed well even when moderate blood volume was withdrawn. The results suggest that the proposed features and algorithm are suitable for the automatic discrimination between hypovolemia and euvolemia, and can be beneficial and applicable in both intraoperative/emergency and combat casualty care.

[1]  Ki H. Chon,et al.  Detection of blood loss in trauma patients using time-frequency analysis of photoplethysmographic signal , 2016, 2016 IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI).

[2]  C. Boer,et al.  Detection of volume loss using the Nexfin device in blood donors , 2016, Anaesthesia.

[3]  Raul Coimbra,et al.  Diagnosis and monitoring of hemorrhagic shock during the initial resuscitation of multiple trauma patients: a review. , 2003, The Journal of emergency medicine.

[4]  Jo Woon Chong,et al.  A Robust Motion Artifact Detection Algorithm for Accurate Detection of Heart Rates From Photoplethysmographic Signals Using Time–Frequency Spectral Features , 2017, IEEE Journal of Biomedical and Health Informatics.

[5]  Jo Woon Chong,et al.  Photoplethysmograph Signal Reconstruction Based on a Novel Hybrid Motion Artifact Detection–Reduction Approach. Part I: Motion and Noise Artifact Detection , 2014, Annals of Biomedical Engineering.

[6]  Ki H. Chon,et al.  Estimation of Respiratory Rate From Photoplethysmogram Data Using Time–Frequency Spectral Estimation , 2009, IEEE Transactions on Biomedical Engineering.

[7]  K. Chon,et al.  A High Resolution Approach to Estimating Time-Frequency Spectra and Their Amplitudes , 2006, Annals of Biomedical Engineering.

[8]  Ki H. Chon,et al.  Multi-channel pulse oximetry for wearable physiological monitoring , 2013, 2013 IEEE International Conference on Body Sensor Networks.

[9]  A. Reisner,et al.  Automated Analysis of Vital Signs to Identify Patients With Substantial Bleeding Before Hospital Arrival: A Feasibility Study , 2015, Shock.

[10]  G. Gutierrez,et al.  Clinical review: Hemorrhagic shock , 2004, Critical care.

[11]  Victor A. Convertino,et al.  Detection of low-volume blood loss: Compensatory reserve versus traditional vital signs , 2014, The journal of trauma and acute care surgery.

[12]  Kathy L Ryan,et al.  Lower body negative pressure as a model to study progression to acute hemorrhagic shock in humans. , 2004, Journal of applied physiology.

[13]  Rolf Lefering,et al.  Impact of hemorrhage on trauma outcome: an overview of epidemiology, clinical presentations, and therapeutic considerations. , 2006, The Journal of trauma.

[14]  J Ludbrook,et al.  Hemodynamic and neurohumoral responses to acute hypovolemia in conscious mammals. , 1991, The American journal of physiology.

[15]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[16]  Radford M. Neal.  Pattern Recognition and Machine Learning , 2007, Technometrics.

[17]  Doreen Eichel,et al.  Learning And Soft Computing Support Vector Machines Neural Networks And Fuzzy Logic Models , 2016 .

[18]  Ki H. Chon,et al.  Improving Pulse Rate Measurements during Random Motion Using a Wearable Multichannel Reflectance Photoplethysmograph , 2016, Sensors.

[19]  G. Grudic,et al.  Estimation of individual-specific progression to impending cardiovascular instability using arterial waveforms. , 2013, Journal of applied physiology.

[20]  S. Shackelford,et al.  Automated prediction of early blood transfusion and mortality in trauma patients , 2014, The journal of trauma and acute care surgery.

[21]  Kirk H. Shelley,et al.  Impact of Withdrawal of 450 ml of Blood on Respiration-Induced Oscillations of the Ear Plethysmographic Waveform , 2007, Journal of Clinical Monitoring and Computing.

[22]  Victor A. Convertino,et al.  Pulse Oximeter Plethysmographic Waveform Changes in Awake, Spontaneously Breathing, Hypovolemic Volunteers , 2011, Anesthesia and analgesia.

[23]  Sylvain Cardin,et al.  Individual-Specific, Beat-to-beat Trending of Significant Human Blood Loss: The Compensatory Reserve , 2015, Shock.

[24]  Jose Salinas,et al.  Heart period variability in trauma patients may predict mortality and allow remote triage. , 2006, Aviation, space, and environmental medicine.

[25]  Ki H. Chon,et al.  A Novel Time-Varying Spectral Filtering Algorithm for Reconstruction of Motion Artifact Corrupted Heart Rate Signals During Intense Physical Activities Using a Wearable Photoplethysmogram Sensor , 2015, Sensors.

[26]  K. Chon,et al.  Using Time-Frequency Analysis of the Photoplethysmographic Waveform to Detect the Withdrawal of 900 mL of Blood , 2012, Anesthesia and analgesia.

[27]  J Ludbrook,et al.  John Ludbrook APPS Symposium Neural Mechanisms In The Cardiovascular Responses To Acute Central Hypovolaemia , 2001, Clinical and experimental pharmacology & physiology.

[28]  J. Kneale.  Understanding hypovolaemic shock , 2003 .

[29]  Jo Woon Chong,et al.  Photoplethysmograph Signal Reconstruction based on a Novel Motion Artifact Detection-Reduction Approach. Part II: Motion and Noise Artifact Removal , 2014, Annals of Biomedical Engineering.

[30]  Kirk H. Shelley,et al.  A Novel Approach Using Time–Frequency Analysis of Pulse-Oximeter Data to Detect Progressive Hypovolemia in Spontaneously Breathing Healthy Subjects , 2011, IEEE Transactions on Biomedical Engineering.