Early Prediction of Hemodynamic Shock in Pediatric Intensive Care Units With Deep Learning on Thermal Videos.
Ontology highlight
ABSTRACT:
SUBMITTER: Vats V
PROVIDER: S-EPMC9340772 | biostudies-literature | 2022
REPOSITORIES: biostudies-literature
Publications
Early Prediction of Hemodynamic Shock in Pediatric Intensive Care Units With Deep Learning on Thermal Videos.
Vats Vanshika V Nagori Aditya A Singh Pradeep P Dutt Raman R Bandhey Harsh H Wason Mahika M Lodha Rakesh R Sethi Tavpritesh T
Frontiers in physiology 20220711
Shock is one of the major killers in intensive care units, and early interventions can potentially reverse it. In this study, we advance a noncontact thermal imaging modality for continuous monitoring of hemodynamic shock working on 1,03,936 frames from 406 videos recorded longitudinally upon 22 pediatric patients. Deep learning was used to preprocess and extract the Center-to-Peripheral Difference (CPD) in temperature values from the videos. This time-series data along with the heart rate was f ...[more]
Similar Datasets
Project description:Hypothermia is a life-threatening condition where the temperature of the body drops below 35°C and is a key source of concern in Intensive Care Units (ICUs). Early identification can help to nudge clinical management to initiate early interventions. Despite its importance, very few studies have focused on the early prediction of hypothermia. In this study, we aim to monitor and predict Hypothermia (30 min-4 h) ahead of its onset using machine learning (ML) models developed on physiological vitals and to prospectively validate the best performing model in the pediatric ICU. We developed and evaluated ML algorithms for the early prediction of hypothermia in a pediatric ICU. Sepsis advanced forecasting engine ICU Database (SafeICU) data resource is an in-house ICU source of data built in the Pediatric ICU at the All-India Institute of Medical Science (AIIMS), New Delhi. Each time-stamp at 1-min resolution was labeled for the presence of hypothermia to construct a retrospective cohort of pediatric patients in the SafeICU data resource. The training set consisted of windows of the length of 4.2 h with a lead time of 30 min-4 h from the onset of hypothermia. A set of 3,835 hand-engineered time-series features were calculated to capture physiological features from the time series. Features selection using the Boruta algorithm was performed to select the most important predictors of hypothermia. A battery of models such as gradient boosting machine, random forest, AdaBoost, and support vector machine (SVM) was evaluated utilizing five-fold test sets. The best-performing model was prospectively validated. A total of 148 patients with 193 ICU stays were eligible for the model development cohort. Of 3,939 features, 726 were statistically significant in the Boruta analysis for the prediction of Hypothermia. The gradient boosting model performed best with an Area Under the Receiver Operating Characteristic curve (AUROC) of 85% (SD = 1.6) and a precision of 59.2% (SD = 8.8) for a 30-min lead time before the onset of Hypothermia onset. As expected, the model showed a decline in model performance at higher lead times, such as AUROC of 77.2% (SD = 2.3) and precision of 41.34% (SD = 4.8) for 4 h ahead of Hypothermia onset. Our GBM(gradient boosting machine) model produced equal and superior results for the prospective validation, where an AUROC of 79.8% and a precision of 53% for a 30-min lead time before the onset of Hypothermia whereas an AUROC of 69.6% and a precision of 38.52% for a (30 min-4 h) lead time prospective validation of Hypothermia. Therefore, this work establishes a pipeline termed ThermoGnose for predicting hypothermia, a major complication in pediatric ICUs.
| S-EPMC9511412 | biostudies-literature
Project description:Proactive detection of hemodynamic shock can prevent organ failure and save lives. Thermal imaging is a non-invasive, non-contact modality to capture body surface temperature with the potential to reveal underlying perfusion disturbance in shock. In this study, we automate early detection and prediction of shock using machine learning upon thermal images obtained in a pediatric intensive care unit of a tertiary care hospital. 539 images were recorded out of which 253 had concomitant measurement of continuous intra-arterial blood pressure, the gold standard for shock monitoring. Histogram of oriented gradient features were used for machine learning based region-of-interest segmentation that achieved 96% agreement with a human expert. The segmented center-to-periphery difference along with pulse rate was used in longitudinal prediction of shock at 0, 3, 6 and 12 hours using a generalized linear mixed-effects model. The model achieved a mean area under the receiver operating characteristic curve of 75% at 0 hours (classification), 77% at 3 hours (prediction) and 69% at 12 hours (prediction) respectively. Since hemodynamic shock associated with critical illness and infectious epidemics such as Dengue is often fatal, our model demonstrates an affordable, non-invasive, non-contact and tele-diagnostic decision support system for its reliable detection and prediction.
| S-EPMC6331545 | biostudies-literature
Project description:BackgroundTimely recognition of hemodynamic instability in critically ill patients enables increased vigilance and early treatment opportunities. We develop the Hemodynamic Stability Index (HSI), which highlights situational awareness of possible hemodynamic instability occurring at the bedside and to prompt assessment for potential hemodynamic interventions.MethodsWe used an ensemble of decision trees to obtain a real-time risk score that predicts the initiation of hemodynamic interventions an hour into the future. We developed the model using the eICU Research Institute (eRI) database, based on adult ICU admissions from 2012 to 2016. A total of 208,375 ICU stays met the inclusion criteria, with 32,896 patients (prevalence = 18%) experiencing at least one instability event where they received one of the interventions during their stay. Predictors included vital signs, laboratory measurements, and ventilation settings.ResultsHSI showed significantly better performance compared to single parameters like systolic blood pressure and shock index (heart rate/systolic blood pressure) and showed good generalization across patient subgroups. HSI AUC was 0.82 and predicted 52% of all hemodynamic interventions with a lead time of 1-h with a specificity of 92%. In addition to predicting future hemodynamic interventions, our model provides confidence intervals and a ranked list of clinical features that contribute to each prediction. Importantly, HSI can use a sparse set of physiologic variables and abstains from making a prediction when the confidence is below an acceptable threshold.ConclusionsThe HSI algorithm provides a single score that summarizes hemodynamic status in real time using multiple physiologic parameters in patient monitors and electronic medical records (EMR). Importantly, HSI is designed for real-world deployment, demonstrating generalizability, strong performance under different data availability conditions, and providing model explanation in the form of feature importance and prediction confidence.
| S-EPMC8590869 | biostudies-literature
Project description:ObjectivesEarly hemorrhage detection in intensive care units (ICUs) enables timely intervention and reduces the risk of irreversible outcomes. In this study, we aimed to develop a machine learning model to predict hemorrhage by learning the patterns of continuously changing, real-world clinical data.MethodsWe used the Medical Information Mart for Intensive Care databases (MIMIC-III and MIMIC-IV). A recurrent neural network was used to predict severe hemorrhage in the ICU. We developed three machine learning models with an increasing number of input features and levels of complexity: model 1 (11 features), model 2 (18 features), and model 3 (27 features). MIMIC-III was used for model training, and MIMIC-IV was split for internal validation. Using the model with the highest performance, external verification was performed using data from a subgroup extracted from the eICU Collaborative Research Database.ResultsWe included 5,670 ICU admissions, with 3,150 in the training set and 2,520 in the internal test set. A positive correlation was found between model complexity and performance. As a measure of performance, three models developed with an increasing number of features showed area under the receiver operating characteristic (AUROC) curve values of 0.61-0.94 according to the range of input data. In the subgroup extracted from the eICU database for external validation, an AUROC value of 0.74 was observed.ConclusionsMachine learning models that rely on real clinical data can be used to predict patients at high risk of bleeding in the ICU.
| S-EPMC9672494 | biostudies-literature
Project description:AimsMyocardial infarction and heart failure are major cardiovascular diseases that affect millions of people in the USA with morbidity and mortality being highest among patients who develop cardiogenic shock. Early recognition of cardiogenic shock allows prompt implementation of treatment measures. Our objective is to develop a new dynamic risk score, called CShock, to improve early detection of cardiogenic shock in the cardiac intensive care unit (ICU).Methods and resultsWe developed and externally validated a deep learning-based risk stratification tool, called CShock, for patients admitted into the cardiac ICU with acute decompensated heart failure and/or myocardial infarction to predict the onset of cardiogenic shock. We prepared a cardiac ICU dataset using the Medical Information Mart for Intensive Care-III database by annotating with physician-adjudicated outcomes. This dataset which consisted of 1500 patients with 204 having cardiogenic/mixed shock was then used to train CShock. The features used to train the model for CShock included patient demographics, cardiac ICU admission diagnoses, routinely measured laboratory values and vital signs, and relevant features manually extracted from echocardiogram and left heart catheterization reports. We externally validated the risk model on the New York University (NYU) Langone Health cardiac ICU database which was also annotated with physician-adjudicated outcomes. The external validation cohort consisted of 131 patients with 25 patients experiencing cardiogenic/mixed shock. CShock achieved an area under the receiver operator characteristic curve (AUROC) of 0.821 (95% CI 0.792-0.850). CShock was externally validated in the more contemporary NYU cohort and achieved an AUROC of 0.800 (95% CI 0.717-0.884), demonstrating its generalizability in other cardiac ICUs. Having an elevated heart rate is most predictive of cardiogenic shock development based on Shapley values. The other top 10 predictors are having an admission diagnosis of myocardial infarction with ST-segment elevation, having an admission diagnosis of acute decompensated heart failure, Braden Scale, Glasgow Coma Scale, blood urea nitrogen, systolic blood pressure, serum chloride, serum sodium, and arterial blood pH.ConclusionThe novel CShock score has the potential to provide automated detection and early warning for cardiogenic shock and improve the outcomes for millions of patients who suffer from myocardial infarction and heart failure.
| S-EPMC11214586 | biostudies-literature
Project description:Purpose: To develop and validate machine learning models for predicting the length of stay (LOS) in the Pediatric Intensive Care Unit (PICU) using data from the Virtual Pediatric Systems (VPS) database. Methods: A retrospective study was conducted utilizing machine learning (ML) algorithms to analyze and predict PICU LOS based on historical patient data from the VPS database. The study included data from over 100 North American PICUs spanning the years 2015-2020. After excluding entries with missing variables and those indicating recovery from cardiac surgery, the dataset comprised 123,354 patient encounters. Various ML models, including Support Vector Machine, Stochastic Gradient Descent Classifier, K-Nearest Neighbors, Decision Tree, Gradient Boosting, CatBoost, and Recurrent Neural Networks (RNNs), were evaluated for their accuracy in predicting PICU LOS at thresholds of 24 h, 36 h, 48 h, 72 h, 5 days, and 7 days. Results: Gradient Boosting, CatBoost, and RNN models demonstrated the highest accuracy, particularly at the 36 h and 48 h thresholds, with accuracy rates between 70 and 73%. These results far outperform traditional statistical and existing prediction methods that report accuracy of only around 50%, which is effectively unusable in the practical setting. These models also exhibited balanced performance between sensitivity (up to 74%) and specificity (up to 82%) at these thresholds. Conclusions: ML models, particularly Gradient Boosting, CatBoost, and RNNs, show moderate effectiveness in predicting PICU LOS with accuracy slightly over 70%, outperforming previously reported human predictions. This suggests potential utility in enhancing resource and staffing management in PICUs. However, further improvements through training on specialized databases can potentially achieve better accuracy and clinical applicability.
| S-EPMC11504172 | biostudies-literature
Project description:This paper targets a major challenge of how to effectively allocate medical resources in intensive care units (ICUs). We trained multiple regression models using the Medical Information Mart for Intensive Care III (MIMIC III) database recorded in the period between 2001 and 2012. The training and validation dataset included pneumonia, sepsis, congestive heart failure, hypotension, chest pain, coronary artery disease, fever, respiratory failure, acute coronary syndrome, shortness of breath, seizure and transient ischemic attack, and aortic stenosis patients' recorded data. Then we tested the models on the unseen data of patients diagnosed with coronary artery disease, congestive heart failure or acute coronary syndrome. We included the admission characteristics, clinical prescriptions, physiological measurements, and discharge characteristics of those patients. We assessed the models' performance using mean residuals and running times as metrics. We ran multiple experiments to study the data partition's impact on the learning phase. The total running time of our best-evaluated model is 123,450.9 mS. The best model gives an average accuracy of 98%, highlighting the location of discharge, initial diagnosis, location of admission, drug therapy, length of stay and internal transfers as the most influencing patterns to decide a patient's readiness for discharge.
| S-EPMC9222879 | biostudies-literature
Project description:Objectives: This study aims to investigate whether the machine learning algorithms could provide an optimal early mortality prediction method compared with other scoring systems for patients with cerebral hemorrhage in intensive care units in clinical practice. Methods: Between 2008 and 2012, from Intensive Care III (MIMIC-III) database, all cerebral hemorrhage patients monitored with the MetaVision system and admitted to intensive care units were enrolled in this study. The calibration, discrimination, and risk classification of predicted hospital mortality based on machine learning algorithms were assessed. The primary outcome was hospital mortality. Model performance was assessed with accuracy and receiver operating characteristic curve analysis. Results: Of 760 cerebral hemorrhage patients enrolled from MIMIC database [mean age, 68.2 years (SD, ±15.5)], 383 (50.4%) patients died in hospital, and 377 (49.6%) patients survived. The area under the receiver operating characteristic curve (AUC) of six machine learning algorithms was 0.600 (nearest neighbors), 0.617 (decision tree), 0.655 (neural net), 0.671(AdaBoost), 0.819 (random forest), and 0.725 (gcForest). The AUC was 0.423 for Acute Physiology and Chronic Health Evaluation II score. The random forest had the highest specificity and accuracy, as well as the greatest AUC, showing the best ability to predict in-hospital mortality. Conclusions: Compared with conventional scoring system and the other five machine learning algorithms in this study, random forest algorithm had better performance in predicting in-hospital mortality for cerebral hemorrhage patients in intensive care units, and thus further research should be conducted on random forest algorithm.
| S-EPMC7855582 | biostudies-literature
Project description:BackgroundEarly identification of children at risk for cardiac arrest would allow for skill training associated with improved outcomes and provides a prevention opportunity.ObjectiveDevelop and assess a predictive model for cardiopulmonary arrest using data available in the first 4 h.MethodsData from PICU patients from 8 institutions included descriptive, severity of illness, cardiac arrest, and outcomes.ResultsOf the 10074 patients, 120 satisfying inclusion criteria sustained a cardiac arrest and 67 (55.9%) died. In univariate analysis, patients with cardiac arrest prior to admission were over 6 times and those with cardiac arrests during the first 4 h were over 50 times more likely to have a subsequent arrest. The multivariate logistic regression model performance was excellent (area under the ROC curve = 0.85 and Hosmer-Lemeshow statistic, p = 0.35). The variables with the highest odds ratio's for sustaining a cardiac arrest in the multivariable model were admission from an inpatient unit (8.23 (CI: 4.35-15.54)), and cardiac arrest in the first 4 h (6.48 (CI: 2.07-20.36). The average risk predicted by the model was highest (11.6%) among children sustaining an arrest during hours >4-12 and continued to be high even for days after the risk assessment period; the average predicted risk was 9.5% for arrests that occurred after 8 PICU days.ConclusionsPatients at high risk of cardiac arrest can be identified with routinely available data after 4 h. The cardiac arrest may occur relatively close to the risk assessment period or days later.
| S-EPMC6258339 | biostudies-literature
Project description:BackgroundLaboratory testing is routinely used to assay blood biomarkers to provide information on physiologic state beyond what clinicians can evaluate from interpreting medical imaging. We hypothesized that deep learning interpretation of echocardiogram videos can provide additional value in understanding disease states and can evaluate common biomarkers results.MethodsWe developed EchoNet-Labs, a video-based deep learning algorithm to detect evidence of anemia, elevated B-type natriuretic peptide (BNP), troponin I, and blood urea nitrogen (BUN), as well as values of ten additional lab tests directly from echocardiograms. We included patients (n = 39,460) aged 18 years or older with one or more apical-4-chamber echocardiogram videos (n = 70,066) from Stanford Healthcare for training and internal testing of EchoNet-Lab's performance in estimating the most proximal biomarker result. Without fine-tuning, the performance of EchoNet-Labs was further evaluated on an additional external test dataset (n = 1,301) from Cedars-Sinai Medical Center. We calculated the area under the curve (AUC) of the receiver operating characteristic curve for the internal and external test datasets.FindingsOn the held-out test set of Stanford patients not previously seen during model training, EchoNet-Labs achieved an AUC of 0.80 (0.79-0.81) in detecting anemia (low hemoglobin), 0.86 (0.85-0.88) in detecting elevated BNP, 0.75 (0.73-0.78) in detecting elevated troponin I, and 0.74 (0.72-0.76) in detecting elevated BUN. On the external test dataset from Cedars-Sinai, EchoNet-Labs achieved an AUC of 0.80 (0.77-0.82) in detecting anemia, of 0.82 (0.79-0.84) in detecting elevated BNP, of 0.75 (0.72-0.78) in detecting elevated troponin I, and of 0.69 (0.66-0.71) in detecting elevated BUN. We further demonstrate the utility of the model in detecting abnormalities in 10 additional lab tests. We investigate the features necessary for EchoNet-Labs to make successful detection and identify potential mechanisms for each biomarker using well-known and novel explainability techniques.InterpretationThese results show that deep learning applied to diagnostic imaging can provide additional clinical value and identify phenotypic information beyond current imaging interpretation methods.FundingJ.W.H. and B.H. are supported by the NSF Graduate Research Fellowship. D.O. is supported by NIH K99 HL157421-01. J.Y.Z. is supported by NSF CAREER 1942926, NIH R21 MD012867-01, NIH P30AG059307 and by a Chan-Zuckerberg Biohub Fellowship.
| S-EPMC8524103 | biostudies-literature