1
|
Rashedi N, Sun Y, Vaze V, Shah P, Halter R, Elliott JT, Paradis NA. Prediction of Occult Hemorrhage in the Lower Body Negative Pressure Model: Initial Validation of Machine Learning Approaches. Mil Med 2024; 189:e1629-e1636. [PMID: 38537150 DOI: 10.1093/milmed/usae061] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/04/2023] [Revised: 10/09/2023] [Accepted: 02/12/2024] [Indexed: 07/05/2024] Open
Abstract
INTRODUCTION Detection of occult hemorrhage (OH) before progression to clinically apparent changes in vital signs remains an important clinical problem in managing trauma patients. The resource-intensiveness associated with continuous clinical patient monitoring and rescue from frank shock makes accurate early detection and prediction with noninvasive measurement technology a desirable innovation. Despite significant efforts directed toward the development of innovative noninvasive diagnostics, the implementation and performance of the newest bedside technologies remain inadequate. This poor performance may reflect the limitations of univariate systems based on one sensor in one anatomic location. It is possible that when signals are measured with multiple modalities in multiple locations, the resulting multivariate anatomic and temporal patterns of measured signals may provide additional discriminative power over single technology univariate measurements. We evaluated the potential superiority of multivariate methods over univariate methods. Additionally, we utilized machine learning-based models to compare the performance of noninvasive-only to noninvasive-plus-invasive measurements in predicting the onset of OH. MATERIALS AND METHODS We applied machine learning methods to preexisting datasets derived using the lower body negative pressure human model of simulated hemorrhage. Employing multivariate measured physiological signals, we investigated the extent to which machine learning methods can effectively predict the onset of OH. In particular, we applied 2 ensemble learning methods, namely, random forest and gradient boosting. RESULTS Analysis of precision, recall, and area under the receiver operating characteristic curve showed a superior performance of multivariate approach to that of the univariate ones. In addition, when using both invasive and noninvasive features, random forest classifier had a recall 95% confidence interval (CI) of 0.81 to 0.86 with a precision 95% CI of 0.65 to 0.72. Interestingly, when only noninvasive features were employed, the results worsened only slightly to a recall 95% CI of 0.80 to 0.85 and a precision 95% CI of 0.61 to 0.73. CONCLUSIONS Multivariate ensemble machine learning-based approaches for the prediction of hemodynamic instability appear to hold promise for the development of effective solutions. In the lower body negative pressure multivariate hemorrhage model, predictions based only on noninvasive measurements performed comparably to those using both invasive and noninvasive measurements.
Collapse
Affiliation(s)
- Navid Rashedi
- Department of Engineering Sciences, Thayer School of Engineering, Dartmouth College, Hanover, NH 03755, USA
| | - Yifei Sun
- Department of Engineering Sciences, Thayer School of Engineering, Dartmouth College, Hanover, NH 03755, USA
| | - Vikrant Vaze
- Department of Engineering Sciences, Thayer School of Engineering, Dartmouth College, Hanover, NH 03755, USA
| | - Parikshit Shah
- Department of Electrical Engineering and Computer Science, Insight Research, Emerald Hills, CA 94065, USA
| | - Ryan Halter
- Department of Engineering Sciences, Thayer School of Engineering, Dartmouth College, Hanover, NH 03755, USA
| | - Jonathan T Elliott
- Department of Emergency Medicine, Geisel School of Medicine, Dartmouth College, Hanover, NH 03755, USA
| | - Norman A Paradis
- Department of Emergency Medicine, Geisel School of Medicine, Dartmouth College, Hanover, NH 03755, USA
| |
Collapse
|
2
|
Webb KL, Pruter WW, Poole RJ, Techentin RW, Johnson CP, Regimbal RJ, Berndt KJ, Holmes DR, Haider CR, Joyner MJ, Convertino VA, Wiggins CC, Curry TB. Comparing the compensatory reserve metric obtained from invasive arterial measurements and photoplethysmographic volume-clamp during simulated hemorrhage. J Clin Monit Comput 2024:10.1007/s10877-024-01166-x. [PMID: 38733507 DOI: 10.1007/s10877-024-01166-x] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/21/2023] [Accepted: 04/16/2024] [Indexed: 05/13/2024]
Abstract
PURPOSE The compensatory reserve metric (CRM) is a novel tool to predict cardiovascular decompensation during hemorrhage. The CRM is traditionally computed using waveforms obtained from photoplethysmographic volume-clamp (PPGVC), yet invasive arterial pressures may be uniquely available. We aimed to examine the level of agreement of CRM values computed from invasive arterial-derived waveforms and values computed from PPGVC-derived waveforms. METHODS Sixty-nine participants underwent graded lower body negative pressure to simulate hemorrhage. Waveform measurements from a brachial arterial catheter and PPGVC finger-cuff were collected. A PPGVC brachial waveform was reconstructed from the PPGVC finger waveform. Thereafter, CRM values were computed using a deep one-dimensional convolutional neural network for each of the following source waveforms; (1) invasive arterial, (2) PPGVC brachial, and (3) PPGVC finger. Bland-Altman analyses were used to determine the level of agreement between invasive arterial CRM values and PPGVC CRM values, with results presented as the Mean Bias [95% Limits of Agreement]. RESULTS The mean bias between invasive arterial- and PPGVC brachial CRM values at rest, an applied pressure of -45mmHg, and at tolerance was 6% [-17%, 29%], 1% [-28%, 30%], and 0% [-25%, 25%], respectively. Additionally, the mean bias between invasive arterial- and PPGVC finger CRM values at rest, applied pressure of -45mmHg, and tolerance was 2% [-22%, 26%], 8% [-19%, 35%], and 5% [-15%, 25%], respectively. CONCLUSION There is generally good agreement between CRM values obtained from invasive arterial waveforms and values obtained from PPGVC waveforms. Invasive arterial waveforms may serve as an alternative for computation of the CRM.
Collapse
Affiliation(s)
- Kevin L Webb
- Department of Anesthesiology and Perioperative Medicine, Mayo Clinic, 200 First St. SW, 55905, Rochester, Minnesota, MN, USA
- Department of Physiology and Biomedical Engineering, Mayo Clinic, Rochester, Minnesota, United States of America
| | - Wyatt W Pruter
- Department of Anesthesiology and Perioperative Medicine, Mayo Clinic, 200 First St. SW, 55905, Rochester, Minnesota, MN, USA
| | - Ruth J Poole
- Special Purpose Processor Development Group, Mayo Clinic, Rochester, Minnesota, United States of America
| | - Robert W Techentin
- Special Purpose Processor Development Group, Mayo Clinic, Rochester, Minnesota, United States of America
| | - Christopher P Johnson
- Department of Anesthesiology and Perioperative Medicine, Mayo Clinic, 200 First St. SW, 55905, Rochester, Minnesota, MN, USA
| | - Riley J Regimbal
- Department of Anesthesiology and Perioperative Medicine, Mayo Clinic, 200 First St. SW, 55905, Rochester, Minnesota, MN, USA
| | - Kaylah J Berndt
- Special Purpose Processor Development Group, Mayo Clinic, Rochester, Minnesota, United States of America
| | - David R Holmes
- Biomedical Analytics and Computational Engineering Laboratory, Mayo Clinic, Rochester, Minnesota, United States of America
| | - Clifton R Haider
- Department of Physiology and Biomedical Engineering, Mayo Clinic, Rochester, Minnesota, United States of America
| | - Michael J Joyner
- Department of Anesthesiology and Perioperative Medicine, Mayo Clinic, 200 First St. SW, 55905, Rochester, Minnesota, MN, USA
| | - Victor A Convertino
- Battlefield Health & Trauma Center for Human Integrative Physiology, Army Institute of Surgical Research, JBSA Fort Sam Houston, San Antonio, TX, United States of America
| | - Chad C Wiggins
- Department of Anesthesiology and Perioperative Medicine, Mayo Clinic, 200 First St. SW, 55905, Rochester, Minnesota, MN, USA
- Department of Kinesiology, Michigan State University, East Lansing, Michigan, United States of America
| | - Timothy B Curry
- Department of Anesthesiology and Perioperative Medicine, Mayo Clinic, 200 First St. SW, 55905, Rochester, Minnesota, MN, USA.
| |
Collapse
|
3
|
Pramanik L, Felton CL, Techentin RW, Holmes DR, Curry TB, Joyner MJ, Convertino VA, Haider CR. Towards a Lightweight Classifier to Detect Hypovolemic Shock. ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY. IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY. ANNUAL INTERNATIONAL CONFERENCE 2023; 2023:1-5. [PMID: 38083358 DOI: 10.1109/embc40787.2023.10340949] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 12/18/2023]
Abstract
Predicting the ability of an individual to compensate for blood loss during hemorrhage and detect the likely onset of hypovolemic shock is necessary to permit early clinical intervention. Towards this end, the compensatory reserve metric (CRM) has been demonstrated to directly correlate with an individual's ability to maintain compensatory mechanisms during loss of blood volume from onset (one-hundred percent health) to exsanguination (zero percent health). This effort describes a lightweight, three-class predictor (good, fair, poor) of an individual's compensatory reserve using a linear support-vector machine (SVM) classifier. A moving mean filter of the predictions demonstrates a feasible model for implementation of real-time hypovolemia monitoring on a wearable device, requiring only 408 bytes to store the models' coefficients and minimal processor cycles to complete the computations.
Collapse
|
4
|
Bedolla CN, Gonzalez JM, Vega SJ, Convertino VA, Snider EJ. An Explainable Machine-Learning Model for Compensatory Reserve Measurement: Methods for Feature Selection and the Effects of Subject Variability. Bioengineering (Basel) 2023; 10:bioengineering10050612. [PMID: 37237682 DOI: 10.3390/bioengineering10050612] [Citation(s) in RCA: 2] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/20/2023] [Revised: 05/08/2023] [Accepted: 05/10/2023] [Indexed: 05/28/2023] Open
Abstract
Tracking vital signs accurately is critical for triaging a patient and ensuring timely therapeutic intervention. The patient's status is often clouded by compensatory mechanisms that can mask injury severity. The compensatory reserve measurement (CRM) is a triaging tool derived from an arterial waveform that has been shown to allow for earlier detection of hemorrhagic shock. However, the deep-learning artificial neural networks developed for its estimation do not explain how specific arterial waveform elements lead to predicting CRM due to the large number of parameters needed to tune these models. Alternatively, we investigate how classical machine-learning models driven by specific features extracted from the arterial waveform can be used to estimate CRM. More than 50 features were extracted from human arterial blood pressure data sets collected during simulated hypovolemic shock resulting from exposure to progressive levels of lower body negative pressure. A bagged decision tree design using the ten most significant features was selected as optimal for CRM estimation. This resulted in an average root mean squared error in all test data of 0.171, similar to the error for a deep-learning CRM algorithm at 0.159. By separating the dataset into sub-groups based on the severity of simulated hypovolemic shock withstood, large subject variability was observed, and the key features identified for these sub-groups differed. This methodology could allow for the identification of unique features and machine-learning models to differentiate individuals with good compensatory mechanisms against hypovolemia from those that might be poor compensators, leading to improved triage of trauma patients and ultimately enhancing military and emergency medicine.
Collapse
Affiliation(s)
- Carlos N Bedolla
- U.S. Army Institute of Surgical Research, JBSA Fort Sam Houston, San Antonio, TX 78234, USA
| | - Jose M Gonzalez
- U.S. Army Institute of Surgical Research, JBSA Fort Sam Houston, San Antonio, TX 78234, USA
| | - Saul J Vega
- U.S. Army Institute of Surgical Research, JBSA Fort Sam Houston, San Antonio, TX 78234, USA
| | - Víctor A Convertino
- U.S. Army Institute of Surgical Research, JBSA Fort Sam Houston, San Antonio, TX 78234, USA
- Department of Medicine, Uniformed Services University, Bethesda, MD 20814, USA
- Department of Emergency Medicine, University of Texas Health, San Antonio, TX 78229, USA
- Department of Biomedical Engineering, University of Texas Health, San Antonio, TX 78249, USA
| | - Eric J Snider
- U.S. Army Institute of Surgical Research, JBSA Fort Sam Houston, San Antonio, TX 78234, USA
| |
Collapse
|
5
|
Gupta JF, Arshad SH, Telfer BA, Snider EJ, Convertino VA. Noninvasive Monitoring of Simulated Hemorrhage and Whole Blood Resuscitation. BIOSENSORS 2022; 12:bios12121168. [PMID: 36551134 PMCID: PMC9775873 DOI: 10.3390/bios12121168] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 10/19/2022] [Revised: 12/03/2022] [Accepted: 12/08/2022] [Indexed: 06/01/2023]
Abstract
Hemorrhage is the leading cause of preventable death from trauma. Accurate monitoring of hemorrhage and resuscitation can significantly reduce mortality and morbidity but remains a challenge due to the low sensitivity of traditional vital signs in detecting blood loss and possible hemorrhagic shock. Vital signs are not reliable early indicators because of physiological mechanisms that compensate for blood loss and thus do not provide an accurate assessment of volume status. As an alternative, machine learning (ML) algorithms that operate on an arterial blood pressure (ABP) waveform have been shown to provide an effective early indicator. However, these ML approaches lack physiological interpretability. In this paper, we evaluate and compare the performance of ML models trained on nine ABP-derived features that provide physiological insight, using a database of 13 human subjects from a lower-body negative pressure (LBNP) model of progressive central hypovolemia and subsequent progressive restoration to normovolemia (i.e., simulated hemorrhage and whole blood resuscitation). Data were acquired at multiple repressurization rates for each subject to simulate varying resuscitation rates, resulting in 52 total LBNP collections. This work is the first to use a single ABP-based algorithm to monitor both simulated hemorrhage and resuscitation. A gradient-boosted regression tree model trained on only the half-rise to dicrotic notch (HRDN) feature achieved a root-mean-square error (RMSE) of 13%, an R2 of 0.82, and area under the receiver operating characteristic curve of 0.97 for detecting decompensation. This single-feature model's performance compares favorably to previously reported results from more-complex black box machine learning models. This model further provides physiological insight because HRDN represents an approximate measure of the delay between the ABP ejected and reflected wave and therefore is an indication of cardiac and peripheral vascular mechanisms that contribute to the compensatory response to blood loss and replacement.
Collapse
Affiliation(s)
- Jay F. Gupta
- Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, MA 02421, USA
| | - Saaid H. Arshad
- Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, MA 02421, USA
| | - Brian A. Telfer
- Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, MA 02421, USA
| | - Eric J. Snider
- U.S. Army Institute of Surgical Research, San Antonio, TX 78234, USA
| | | |
Collapse
|
6
|
Gupta JF, Telfer BA, Convertino VA. Feature Importance Analysis for Compensatory Reserve to Predict Hemorrhagic Shock. ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY. IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY. ANNUAL INTERNATIONAL CONFERENCE 2022; 2022:1747-1752. [PMID: 36086009 DOI: 10.1109/embc48229.2022.9871661] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/15/2023]
Abstract
Hemorrhage is the leading cause of preventable death from trauma. Traditionally, vital signs have been used to detect blood loss and possible hemorrhagic shock. However, vital signs are not sensitive for early detection because of physiological mechanisms that compensate for blood loss. As an alternative, machine learning algorithms that operate on an arterial blood pressure (ABP) waveform acquired via photoplethysmography have been shown to provide an effective early indicator. However, these machine learning approaches lack physiological interpretability. In this paper, we evaluate the importance of nine ABP-derived features that provide physiological insight, using a database of 40 human subjects from a lower-body negative pressure model of progressive central hypovolemia. One feature was found to be considerably more important than any other. That feature, the half-rise to dicrotic notch (HRDN), measures an approximate time delay between the ABP ejected and reflected wave components. This delay is an indication of compensatory mechanisms such as reduced arterial compliance and vasoconstriction. For a scale of 0% to 100%, with 100% representing normovolemia and 0% representing decompensation, linear regression of the HRDN feature results in root-mean-squared error of 16.9%, R2 of 0.72, and an area under the receiver operating curve for detecting decompensation of 0.88. These results are comparable to previously reported results from the more complex black box machine learning models. Clinical Relevance- A single physiologically interpretable feature measured from an arterial blood pressure waveform is shown to be effective in monitoring for blood loss and impending hemorrhagic shock based on data from a human lower-body negative pressure model of progressive central hypolemia.
Collapse
|
7
|
AI-Enabled Advanced Development for Assessing Low Circulating Blood Volume for Emergency Medical Care: Comparison of Compensatory Reserve Machine-Learning Algorithms. SENSORS 2022; 22:s22072642. [PMID: 35408255 PMCID: PMC9003258 DOI: 10.3390/s22072642] [Citation(s) in RCA: 4] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 12/10/2021] [Revised: 03/22/2022] [Accepted: 03/24/2022] [Indexed: 11/21/2022]
Abstract
The application of artificial intelligence (AI) has provided new capabilities to develop advanced medical monitoring sensors for detection of clinical conditions of low circulating blood volume such as hemorrhage. The purpose of this study was to compare for the first time the discriminative ability of two machine learning (ML) algorithms based on real-time feature analysis of arterial waveforms obtained from a non-invasive continuous blood pressure system (Finometer®) signal to predict the onset of decompensated shock: the compensatory reserve index (CRI) and the compensatory reserve metric (CRM). One hundred ninety-one healthy volunteers underwent progressive simulated hemorrhage using lower body negative pressure (LBNP). The least squares means and standard deviations for each measure were assessed by LBNP level and stratified by tolerance status (high vs. low tolerance to central hypovolemia). Generalized Linear Mixed Models were used to perform repeated measures logistic regression analysis by regressing the onset of decompensated shock on CRI and CRM. Sensitivity and specificity were assessed by calculation of receiver-operating characteristic (ROC) area under the curve (AUC) for CRI and CRM. Values for CRI and CRM were not distinguishable across levels of LBNP independent of LBNP tolerance classification, with CRM ROC AUC (0.9268) being statistically similar (p = 0.134) to CRI ROC AUC (0.9164). Both CRI and CRM ML algorithms displayed discriminative ability to predict decompensated shock to include individual subjects with varying levels of tolerance to central hypovolemia. Arterial waveform feature analysis provides a highly sensitive and specific monitoring approach for the detection of ongoing hemorrhage, particularly for those patients at greatest risk for early onset of decompensated shock and requirement for implementation of life-saving interventions.
Collapse
|