Comparison of machine learning algorithms for mortality prediction in intensive care patients on multi-center critical care databases

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.

Download Full-text

Patient Survival Prediction with Machine Learning Algorithms

Journal of Intelligent Systems with Applications ◽

10.54856/jiswa.202012126 ◽

2020 ◽

pp. 93-96

Author(s):

Mustafa Berkant Selek ◽

Saadet Sena Egeli ◽

Yalcin Isler

Keyword(s):

Machine Learning ◽

Intensive Care ◽

Cross Validation ◽

Patient Survival ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Survival Prediction ◽

Accuracy Rate ◽

Intensive Care Patients ◽

One Year

In this study, the intensive care unit patient survival is predicted by machine learning algorithms according to the examinations performed in the first 24 hours. The data of intensive care patients collected from approximately two hundred hospitals over a period of one year were used. Algorithms are run in Python environment. Machine learning models were compared with the Cross-Validation method, and the random forest algorithm is used. The model made the prediction with 92,53% accuracy rate.

Download Full-text

Prediction of Mortality in Surgical Intensive Care Unit Patients Using Machine Learning Algorithms

Frontiers in Medicine ◽

10.3389/fmed.2021.621861 ◽

2021 ◽

Vol 8 ◽

Author(s):

Kyongsik Yun ◽

Jihoon Oh ◽

Tae Ho Hong ◽

Eun Young Kim

Keyword(s):

Machine Learning ◽

Intensive Care Unit ◽

Intensive Care ◽

Decision Tree ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Receiver Operating Curve ◽

Surgical Intensive Care Unit ◽

Hospital Death ◽

Surgical Intensive Care

Objective: Predicting prognosis of in-hospital patients is critical. However, it is challenging to accurately predict the life and death of certain patients at certain period. To determine whether machine learning algorithms could predict in-hospital death of critically ill patients with considerable accuracy and identify factors contributing to the prediction power.Materials and Methods: Using medical data of 1,384 patients admitted to the Surgical Intensive Care Unit (SICU) of our institution, we investigated whether machine learning algorithms could predict in-hospital death using demographic, laboratory, and other disease-related variables, and compared predictions using three different algorithmic methods. The outcome measurement was the incidence of unexpected postoperative mortality which was defined as mortality without pre-existing not-for-resuscitation order that occurred within 30 days of the surgery or within the same hospital stay as the surgery.Results: Machine learning algorithms trained with 43 variables successfully classified dead and live patients with very high accuracy. Most notably, the decision tree showed the higher classification results (Area Under the Receiver Operating Curve, AUC = 0.96) than the neural network classifier (AUC = 0.80). Further analysis provided the insight that serum albumin concentration, total prenatal nutritional intake, and peak dose of dopamine drug played an important role in predicting the mortality of SICU patients.Conclusion: Our results suggest that machine learning algorithms, especially the decision tree method, can provide information on structured and explainable decision flow and accurately predict hospital mortality in SICU hospitalized patients.

Download Full-text

Early diagnosis of bloodstream infections in the intensive care unit using machine-learning algorithms

Intensive Care Medicine ◽

10.1007/s00134-019-05876-8 ◽

2020 ◽

Vol 46 (3) ◽

pp. 454-462 ◽

Cited By ~ 4

Author(s):

Michael Roimi ◽

Ami Neuberger ◽

Anat Shrot ◽

Mical Paul ◽

Yuval Geffen ◽

...

Keyword(s):

Machine Learning ◽

Intensive Care Unit ◽

Intensive Care ◽

Early Diagnosis ◽

Learning Algorithms ◽

Bloodstream Infections ◽

Machine Learning Algorithms

Download Full-text

Abstract 13455: Predicting Emergency Department Disposition at Triage for Suspected Patients With Acute Coronary Syndrome Using Machine Learning Algorithms

Circulation ◽

10.1161/circ.142.suppl_3.13455 ◽

2020 ◽

Vol 142 (Suppl_3) ◽

Author(s):

Stephanie O Frisch ◽

Zeineb Bouzid ◽

Jessica Zègre-Hemsey ◽

Clifton W CALLAWAY ◽

Holli A Devon ◽

...

Keyword(s):

Machine Learning ◽

Acute Coronary Syndrome ◽

Critical Care ◽

Random Forest ◽

Characteristic Curve ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Gradient Boosting ◽

Coronary Syndrome ◽

Critical Care Admission

Introduction: Overcrowded emergency departments (ED) and undifferentiated patients make the provision of care and resources challenging. We examined whether machine learning algorithms could identify ED patients’ disposition (hospitalization and critical care admission) using readily available objective triage data among patients with symptoms suggestive of acute coronary syndrome (ACS). Methods: This was a retrospective observational cohort study of adult patients who were triaged at the ED for a suspected coronary event. A total of 162 input variables (k) were extracted from the electronic health record: demographics (k=3), mode of transportation (k=1), past medical/surgical history (k=57), first ED vital signs (k=7), home medications (k=31), symptomology (k=40), and the computer generated automatic interpretation of 12-lead electrocardiogram (k=23). The primary outcomes were hospitalization and critical care admission (i.e., admission to intensive or step-down care unit). We used 10-fold stratified cross validation to evaluate the performance of five machine learning algorithms to predict the study outcomes: logistic regression, naïve Bayes, random forest, gradient boosting and artificial neural network classifiers. We determined the best model by comparing the area under the receiver operating characteristic curve (AUC) of all models. Results: Included were 1201 patients (age 64±14, 39% female; 10% Black) with a total of 956 hospitalizations, and 169 critical care admissions. The best performing machine learning classifier for the outcome of hospitalization was gradient boosting machine with an AUC of 0.85 (95% CI, 0.82–0.89), 89% sensitivity, and F-score of 0.83; random forest classifier performed the best for the outcome of critical care admission with an AUC of 0.73 (95% CI, 0.70–0.77), 76% sensitivity, and F-score of 0.56. Conclusion: Predictive machine learning algorithms demonstrate excellent to good discriminative power to predict hospitalization and critical care admission, respectively. Administrators and clinicians could benefit from machine learning approaches to predict hospitalization and critical care admission, to optimize and allocate scarce ED and hospital resources and provide optimal care.

Download Full-text

Machine learning algorithms to predict weight gain at discharge in neonatal intensive care unit: state of the art

Clinical Nutrition ESPEN ◽

10.1016/j.clnesp.2021.09.512 ◽

2021 ◽

Vol 46 ◽

pp. S724-S725

Author(s):

N. Yalcin ◽

H.T. Çelik ◽

K. Demirkan ◽

S. Yiğit

Keyword(s):

Machine Learning ◽

Intensive Care Unit ◽

Intensive Care ◽

Weight Gain ◽

Neonatal Intensive Care Unit ◽

Neonatal Intensive Care ◽

State Of The Art ◽

Learning Algorithms ◽

Machine Learning Algorithms

Download Full-text

Using Machine Learning Techniques to Aid Empirical Antibiotic Therapy Decisions in the Intensive Care Unit of a General Hospital in Greece

Antibiotics ◽

10.3390/antibiotics9020050 ◽

2020 ◽

Vol 9 (2) ◽

pp. 50 ◽

Cited By ~ 4

Author(s):

Georgios Feretzakis ◽

Evangelos Loupelis ◽

Aikaterini Sakagianni ◽

Dimitris Kalles ◽

Maria Martsoukou ◽

...

Keyword(s):

Machine Learning ◽

Intensive Care Unit ◽

Intensive Care ◽

Antibiotic Therapy ◽

Antibiotic Susceptibility ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

High Morbidity ◽

Empirical Antibiotic Therapy ◽

Gram Negative

Hospital-acquired infections, particularly in the critical care setting, have become increasingly common during the last decade, with Gram-negative bacterial infections presenting the highest incidence among them. Multi-drug-resistant (MDR) Gram-negative infections are associated with high morbidity and mortality with significant direct and indirect costs resulting from long hospitalization due to antibiotic failure. Time is critical to identifying bacteria and their resistance to antibiotics due to the critical health status of patients in the intensive care unit (ICU). As common antibiotic resistance tests require more than 24 h after the sample is collected to determine sensitivity in specific antibiotics, we suggest applying machine learning (ML) techniques to assist the clinician in determining whether bacteria are resistant to individual antimicrobials by knowing only a sample’s Gram stain, site of infection, and patient demographics. In our single center study, we compared the performance of eight machine learning algorithms to assess antibiotic susceptibility predictions. The demographic characteristics of the patients are considered for this study, as well as data from cultures and susceptibility testing. Applying machine learning algorithms to patient antimicrobial susceptibility data, readily available, solely from the Microbiology Laboratory without any of the patient’s clinical data, even in resource-limited hospital settings, can provide informative antibiotic susceptibility predictions to aid clinicians in selecting appropriate empirical antibiotic therapy. These strategies, when used as a decision support tool, have the potential to improve empiric therapy selection and reduce the antimicrobial resistance burden.

Download Full-text

Machine Learning Algorithms to Predict Mortality of Neonates on Mechanical Intubation for Respiratory Failure

Biomedicines ◽

10.3390/biomedicines9101377 ◽

2021 ◽

Vol 9 (10) ◽

pp. 1377

Author(s):

Jen-Fu Hsu ◽

Chi Yang ◽

Chun-Yuan Lin ◽

Shih-Ming Chu ◽

Hsuan-Rong Huang ◽

...

Keyword(s):

Machine Learning ◽

Respiratory Failure ◽

Learning Algorithms ◽

Illness Severity ◽

Mortality Prediction ◽

Machine Learning Algorithms ◽

Classification And Regression Tree ◽

Net Benefit ◽

Tree Model ◽

Severity Scores

Background: Early identification of critically ill neonates with poor outcomes can optimize therapeutic strategies. We aimed to examine whether machine learning (ML) methods can improve mortality prediction for neonatal intensive care unit (NICU) patients on intubation for respiratory failure. Methods: A total of 1734 neonates with respiratory failure were randomly divided into training (70%, n = 1214) and test (30%, n = 520) sets. The primary outcome was the probability of NICU mortality. The areas under the receiver operating characteristic curves (AUCs) of several ML algorithms were compared with those of the conventional neonatal illness severity scoring systems including the NTISS and SNAPPE-II. Results: For NICU mortality, the random forest (RF) model showed the highest AUC (0.939 (0.921–0.958)) for the prediction of neonates with respiratory failure, and the bagged classification and regression tree model demonstrated the next best results (0.915 (0.891–0.939)). The AUCs of both models were significantly better than the traditional NTISS (0.836 (0.800–0.871)) and SNAPPE-II scores (0.805 (0.766–0.843)). The superior performances were confirmed by higher accuracy and F1 score and better calibration, and the superior and net benefit was confirmed by decision curve analysis. In addition, Shapley additive explanation (SHAP) values were utilized to explain the RF prediction model. Conclusions: Machine learning algorithms increase the accuracy and predictive ability for mortality of neonates with respiratory failure compared with conventional neonatal illness severity scores. The RF model is suitable for clinical use in the NICU, and clinicians can gain insights and have better communication with families in advance.

Download Full-text

Beatmungsassoziierte Pneumonie: Elektronische Nase und künstliche Intelligenz als Potenzial für die Zukunft

Karger Kompass Pneumologie ◽

10.1159/000510425 ◽

2020 ◽

Vol 8 (5) ◽

pp. 254-255

Author(s):

Johannes Knoch

Keyword(s):

Artificial Intelligence ◽

Machine Learning ◽

Intensive Care Unit ◽

Intensive Care ◽

Predictive Value ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Machine Learning Technique ◽

Learning Technique ◽

The Mean

Background: Ventilator-associated pneumonia (VAP) is a significant cause of mortality in the intensive care unit. Early diagnosis of VAP is important to provide appropriate treatment and reduce mortality. Developing a noninvasive and highly accurate diagnostic method is important. The invention of electronic sensors has been applied to analyze the volatile organic compounds in breath to detect VAP using a machine learning technique. However, the process of building an algorithm is usually unclear and prevents physicians from applying the artificial intelligence technique in clinical practice. Clear processes of model building and assessing accuracy are warranted. The objective of this study was to develop a breath test for VAP with a standardized protocol for a machine learning technique. Methods: We conducted a case-control study. This study enrolled subjects in an intensive care unit of a hospital in southern Taiwan from February 2017 to June 2019. We recruited patients with VAP as the case group and ventilated patients without pneumonia as the control group. We collected exhaled breath and analyzed the electric resistance changes of 32 sensor arrays of an electronic nose. We split the data into a set for training algorithms and a set for testing. We applied eight machine learning algorithms to build prediction models, improving model performance and providing an estimated diagnostic accuracy. Results: A total of 33 cases and 26 controls were used in the final analysis. Using eight machine learning algorithms, the mean accuracy in the testing set was 0.81 ± 0.04, the sensitivity was 0.79 ± 0.08, the specificity was 0.83 ± 0.00, the positive predictive value was 0.85 ± 0.02, the negative predictive value was 0.77 ± 0.06, and the area under the receiver operator characteristic curves was 0.85 ± 0.04. The mean kappa value in the testing set was 0.62 ± 0.08, which suggested good agreement. Conclusions: There was good accuracy in detecting VAP by sensor array and machine learning techniques. Artificial intelligence has the potential to assist the physician in making a clinical diagnosis. Clear protocols for data processing and the modeling procedure needed to increase generalizability.

Download Full-text

A Time-Incorporated SOFA-Based Explainable Machine Learning Model for Mortality Prediction in Critically Ill Patients

10.21203/rs.3.rs-1026029/v1 ◽

2021 ◽

Author(s):

Yang Liu ◽

Kun Gao ◽

Hongbin Deng ◽

Tong Ling ◽

Jiajia Lin ◽

...

Keyword(s):

Machine Learning ◽

Intensive Care ◽

Sofa Score ◽

Model Development ◽

Mortality Prediction ◽

Machine Learning Algorithms ◽

Time Dimension ◽

Icu Mortality ◽

Test Set ◽

Individual Scores

Abstract Background: Organ dysfunction (OD) assessment is essential in intensive care units (ICUs). However, no OD scoring system has so far considered the duration of OD, which is clinically relevant. This study aimed to develop and validate an ICU mortality prediction model based on the Sequential Organ Failure Assessment (SOFA) score, incorporating the time dimension with machine learning methods.Methods: Data from the eICU Collaborative Research Database and Medical Information Mart for Intensive Care (MIMIC) -III were mixed for model development, and the MIMIC-IV dataset and Nanjing Jinling Hospital Surgery ICU (SICU-JL) dataset were used for external testing. Adult patients in the ICUs for more than 72 hours were deemed eligible. The total SOFA score and individual scores were calculated every 12 hours for the first three days of ICU admission. Time-dimensional variables were derived from the consecutively recorded SOFA scores and individual scores for each organ. A modified SOFA model incorporating the time dimension (T-SOFA) was stepwise constructed to predict ICU mortality using multiple machine learning algorithms. The predictive performance was assessed with the area under the receiver operating characteristic curves (AUROC). Also, we utilized the SHapley Additive exPlanations (SHAP) algorithm for data visualization and model explainability.Results: We extracted a total of 66,709 ICU patients from the mixed datasets for model development and 15,423 patients for validation. The T-SOFA M3 that incorporated the time dimension features and age, using the XGBoost algorithm, significantly outperformed the original SOFA scores (AUROC 0.800 95% CI [0.787-0.813] vs. 0.693 95% CI [0.678-0.709], p<0.01) in the validation set. Good prediction performance was maintained for the T-SOFA M3 in test Set A and test Set B, with AUROC of 0.803, 95% CI[0.791-0.815], and 0.830, 95%CI [0.789-0.870], respectively. Significant contributors demonstrated by the SHAP analysis included total SOFA score, Respiration-score, CNS-score, age, Cardiovascular-score, and SOFA Organ dysfunction Unalleviated Time Index.Conclusions: A SOFA-based, time-incorporated prediction model was developed and validated by machine learning algorithms, showing satisfactory predictability and medical interpretability.

Download Full-text