Machine learning to predict mortality after rehabilitation among patients with severe stroke

AbstractStroke is among the leading causes of death and disability worldwide. Approximately 20–25% of stroke survivors present severe disability, which is associated with increased mortality risk. Prognostication is inherent in the process of clinical decision-making. Machine learning (ML) methods have gained increasing popularity in the setting of biomedical research. The aim of this study was twofold: assessing the performance of ML tree-based algorithms for predicting three-year mortality model in 1207 stroke patients with severe disability who completed rehabilitation and comparing the performance of ML algorithms to that of a standard logistic regression. The logistic regression model achieved an area under the Receiver Operating Characteristics curve (AUC) of 0.745 and was well calibrated. At the optimal risk threshold, the model had an accuracy of 75.7%, a positive predictive value (PPV) of 33.9%, and a negative predictive value (NPV) of 91.0%. The ML algorithm outperformed the logistic regression model through the implementation of synthetic minority oversampling technique and the Random Forests, achieving an AUC of 0.928 and an accuracy of 86.3%. The PPV was 84.6% and the NPV 87.5%. This study introduced a step forward in the creation of standardisable tools for predicting health outcomes in individuals affected by stroke.

Download Full-text

Predicting the risk of inpatient hypoglycemia with machine learning using electronic health records

10.2337/figshare.12091953.v1 ◽

2020 ◽

Author(s):

Yue Ruan ◽

Alexis Bellot ◽

Zuzana Moysova ◽

Garry D. Tan ◽

Alistair Lumb ◽

...

Keyword(s):

Machine Learning ◽

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model ◽

Prediction Models ◽

Machine Learning Algorithms ◽

Operating Characteristics ◽

Oxygen Saturation Level ◽

Electronic Health ◽

Clinically Significant

Objective We analyzed data from inpatients with diabetes admitted to a large university hospital to predict the risk of hypoglycemia through the use of machine learning algorithms. Research Design and Methods Four years of data was extracted from a hospital electronic health record system. This included laboratory and point-of-care blood glucose (BG) values to identify biochemical and clinically significant hypoglycaemic episodes (BG < 3.9 and < 2.9mmol/L respectively). We used patient demographics, administered medications, vital signs, laboratory results and procedures performed during the hospital stays to inform the model. Two iterations of the dataset included the doses of insulin administered and the past history of inpatient hypoglycaemia. Eighteen different prediction models were compared using the area under curve of the receiver operating characteristics (AUC_ROC) through a ten-fold cross validation. Results We analyzed data obtained from 17,658 inpatients with diabetes who underwent 32,758 admissions between July 2014 and August 2018. The predictive factors from the logistic regression model included people undergoing procedures, weight, type of diabetes, oxygen saturation level, use of medications (insulin, sulfonylurea, metformin) and albumin levels. The machine learning model with the best performance was the XGBoost model (AUC_ROC 0.96. This outperformed the logistic regression model which had an AUC_ROC of 0.75 for the estimation of the risk of clinically significant hypoglycaemia. Conclusions Advanced machine learning models are superior to logistic regression models in predicting the risk of hypoglycemia in inpatients with diabetes. Trials of such models should be conducted in real time to evaluate their utility to reduce inpatient hypoglycaemia.

Download Full-text

Predicting the risk of inpatient hypoglycemia with machine learning using electronic health records

10.2337/figshare.12091953 ◽

2020 ◽

Author(s):

Yue Ruan ◽

Alexis Bellot ◽

Zuzana Moysova ◽

Garry D. Tan ◽

Alistair Lumb ◽

...

Keyword(s):

Machine Learning ◽

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model ◽

Prediction Models ◽

Machine Learning Algorithms ◽

Operating Characteristics ◽

Oxygen Saturation Level ◽

Electronic Health ◽

Clinically Significant

Download Full-text

Logistic Regression Model for Loan Prediction: A Machine Learning Approach

10.1109/eti4.051663.2021.9619201 ◽

2021 ◽

Author(s):

Richa Manglani ◽

Anuja Bokhare

Keyword(s):

Machine Learning ◽

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model ◽

Learning Approach ◽

Machine Learning Approach

Download Full-text

Machine-Learning vs. Expert-Opinion Driven Logistic Regression Modelling for Predicting 30-Day Unplanned Rehospitalisation in Preterm Babies: A Prospective, Population-Based Study (EPIPAGE 2)

Frontiers in Pediatrics ◽

10.3389/fped.2020.585868 ◽

2021 ◽

Vol 8 ◽

Author(s):

Robert A. Reed ◽

Andrei S. Morgan ◽

Jennifer Zeitlin ◽

Pierre-Henri Jarreau ◽

Héloïse Torchin ◽

...

Keyword(s):

Machine Learning ◽

Logistic Regression ◽

Random Forest ◽

Regression Model ◽

Expert Opinion ◽

Logistic Regression Model ◽

Population Based ◽

Regression Modelling ◽

Preterm Babies ◽

Logistic Regression Modelling

Introduction: Preterm babies are a vulnerable population that experience significant short and long-term morbidity. Rehospitalisations constitute an important, potentially modifiable adverse event in this population. Improving the ability of clinicians to identify those patients at the greatest risk of rehospitalisation has the potential to improve outcomes and reduce costs. Machine-learning algorithms can provide potentially advantageous methods of prediction compared to conventional approaches like logistic regression.Objective: To compare two machine-learning methods (least absolute shrinkage and selection operator (LASSO) and random forest) to expert-opinion driven logistic regression modelling for predicting unplanned rehospitalisation within 30 days in a large French cohort of preterm babies.Design, Setting and Participants: This study used data derived exclusively from the population-based prospective cohort study of French preterm babies, EPIPAGE 2. Only those babies discharged home alive and whose parents completed the 1-year survey were eligible for inclusion in our study. All predictive models used a binary outcome, denoting a baby's status for an unplanned rehospitalisation within 30 days of discharge. Predictors included those quantifying clinical, treatment, maternal and socio-demographic factors. The predictive abilities of models constructed using LASSO and random forest algorithms were compared with a traditional logistic regression model. The logistic regression model comprised 10 predictors, selected by expert clinicians, while the LASSO and random forest included 75 predictors. Performance measures were derived using 10-fold cross-validation. Performance was quantified using area under the receiver operator characteristic curve, sensitivity, specificity, Tjur's coefficient of determination and calibration measures.Results: The rate of 30-day unplanned rehospitalisation in the eligible population used to construct the models was 9.1% (95% CI 8.2–10.1) (350/3,841). The random forest model demonstrated both an improved AUROC (0.65; 95% CI 0.59–0.7; p = 0.03) and specificity vs. logistic regression (AUROC 0.57; 95% CI 0.51–0.62, p = 0.04). The LASSO performed similarly (AUROC 0.59; 95% CI 0.53–0.65; p = 0.68) to logistic regression.Conclusions: Compared to an expert-specified logistic regression model, random forest offered improved prediction of 30-day unplanned rehospitalisation in preterm babies. However, all models offered relatively low levels of predictive ability, regardless of modelling method.

Download Full-text

Pure-Tone Hearing Asymmetry: A Logistic Approach Modeling Age, Sex, and Noise Exposure History

Journal of the American Academy of Audiology ◽

10.3766/jaaa.23.7.8 ◽

2012 ◽

Vol 23 (07) ◽

pp. 553-570 ◽

Cited By ~ 3

Author(s):

David A. Zapala ◽

Robin E. Criter ◽

Jamie M. Bogle ◽

Larry B. Lundy ◽

Michael J. Cevette ◽

...

Keyword(s):

Hearing Loss ◽

Logistic Regression ◽

Regression Model ◽

Pure Tone ◽

Logistic Regression Model ◽

Noise Exposure ◽

Clinical Decision Making ◽

Clinical Impression ◽

Patient History ◽

Exposure History

Background: Asymmetric hearing loss (AHL) can be an early sign of vestibular schwannoma (VS). However, recognizing VS-induced AHL is challenging. There is no universally accepted definition of a “medically significant pure-tone hearing asymmetry,” in part because AHL is a common feature of medically benign forms of hearing loss (e.g., age- or firearm-related hearing loss). In most cases, the determination that an observed AHL does not come from a benign cause involves subjective clinical judgment. Purpose: Our purpose was threefold: (1) to quantify hearing asymmetry distributions in a large group of patients with medically benign forms of hearing loss, stratifying for age, sex, and noise exposure history; (2) to assess how previously proposed hearing asymmetry calculations segregate tumor from nontumor cases; and (3) to present the results of a logistic regression method for defining hearing asymmetry that incorporates age, sex, and noise information. Research Design: Retrospective chart review. Study Sample: Five thousand six hundred and sixty-one patients with idiopathic, age- or noise exposure-related hearing loss and 85 untreated VS patients. Data Collection and Analysis: Audiometric, patient history, and clinical impression data were collected from 22,785 consecutive patient visits to the audiology section at Mayo Clinic in Florida from 2006 to 2009 to screen for eligibility. Those eligible were then stratified by VS presence, age, sex, and self-reported noise exposure history. Pure-tone asymmetry distributions were analyzed. Audiometric data from VS diagnoses were used to create four additional audiograms per patient to model the hypothetical development of AHL prior to the actual hearing test. The ability of 11 previously defined hearing asymmetry calculations to distinguish between VS and non-VS cases was described. A logistic regression model was developed that integrated age, sex, and noise exposure history with pure-tone asymmetry data. Regression model performance was then compared to existing asymmetry calculation methods. Results: The 11 existing pure-tone asymmetry calculations varied in tumor detection performance. Age, sex, and noise exposure history helped to predict benign forms of hearing asymmetry. The logistic regression model outperformed existing asymmetry calculations and better accounted for normal age-, sex-, and noise exposure-related asymmetry variability. Conclusions: Our logistic regression asymmetry method improves the clinician's ability to estimate risk of VS, in part by integrating categorical patient history and numeric test data. This form of modeling can enhance clinical decision making in audiology and otology.

Download Full-text

Prediction of perinatal death using machine learning models: a birth registry-based cohort study in northern Tanzania

BMJ Open ◽

10.1136/bmjopen-2020-040132 ◽

2020 ◽

Vol 10 (10) ◽

pp. e040132

Author(s):

Innocent B Mboya ◽

Michael J Mahande ◽

Mohanad Mohammed ◽

Joseph Obure ◽

Henry G Mwambi

Keyword(s):

Machine Learning ◽

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model ◽

Perinatal Death ◽

Learning Models ◽

Net Benefit ◽

Birth Registry ◽

Perinatal Deaths ◽

Machine Learning Models

ObjectiveWe aimed to determine the key predictors of perinatal deaths using machine learning models compared with the logistic regression model.DesignA secondary data analysis using the Kilimanjaro Christian Medical Centre (KCMC) Medical Birth Registry cohort from 2000 to 2015. We assessed the discriminative ability of models using the area under the receiver operating characteristics curve (AUC) and the net benefit using decision curve analysis.SettingThe KCMC is a zonal referral hospital located in Moshi Municipality, Kilimanjaro region, Northern Tanzania. The Medical Birth Registry is within the hospital grounds at the Reproductive and Child Health Centre.ParticipantsSingleton deliveries (n=42 319) with complete records from 2000 to 2015.Primary outcome measuresPerinatal death (composite of stillbirths and early neonatal deaths). These outcomes were only captured before mothers were discharged from the hospital.ResultsThe proportion of perinatal deaths was 3.7%. There were no statistically significant differences in the predictive performance of four machine learning models except for bagging, which had a significantly lower performance (AUC 0.76, 95% CI 0.74 to 0.79, p=0.006) compared with the logistic regression model (AUC 0.78, 95% CI 0.76 to 0.81). However, in the decision curve analysis, the machine learning models had a higher net benefit (ie, the correct classification of perinatal deaths considering a trade-off between false-negatives and false-positives)—over the logistic regression model across a range of threshold probability values.ConclusionsIn this cohort, there was no significant difference in the prediction of perinatal deaths between machine learning and logistic regression models, except for bagging. The machine learning models had a higher net benefit, as its predictive ability of perinatal death was considerably superior over the logistic regression model. The machine learning models, as demonstrated by our study, can be used to improve the prediction of perinatal deaths and triage for women at risk.

Download Full-text

P5710Clinical applications of machine learning for prediction of incident atrial fibrillation from the general population: a nationwide cohort study

European Heart Journal ◽

10.1093/eurheartj/ehz746.0651 ◽

2019 ◽

Vol 40 (Supplement_1) ◽

Author(s):

I.-S Kim ◽

P S Yang ◽

H T Yu ◽

T H Kim ◽

J S Uhm ◽

...

Keyword(s):

Machine Learning ◽

Atrial Fibrillation ◽

Logistic Regression ◽

General Population ◽

Regression Model ◽

National Health ◽

Logistic Regression Model ◽

Learning System ◽

Health Examination ◽

Clinical Variables

Abstract Background To evaluate the ability of machine learning algorithms to predict incident atrial fibrillation (AF) from the general population using health examination items. Methods We included 483,343 subjects who received national health examinations from the Korean National Health Insurance Service-based National Sample Cohort (NHIS-NSC). We trained deep neural network model (DNN) of a deep learning system and decision tree model (DT) of a machine learning system using clinical variables and health examination items (including age, sex, body mass index, history of heart failure, hypertension or diabetes, baseline creatinine, and smoking and alcohol intake habits) to predict incident AF using a training dataset of 341,771 subjects constructed from the NHIS-NSC database. The DNN and DT were validated using an independent test dataset of 141,572 remaining subjects. C-indices of DNN and DT for prediction of incident AF were compared with that of conventional logistic regression model. Results During 1,874,789 person·years (mean±standard-deviation age 47.7±14.4 years, 49.6% male), 3,282 subjects with incident AF were observed. In the validation dataset, 1,139 subjects with incident AF were observed. The c-indices of the DNN and DT for incident AF prediction were 0.828 [0.819–0.836] and 0.835 [0.825–0.844], and were significantly higher (p<0.01) than conventional logistic regression model (c-index=0.789 [0.784–0.794]). Conclusions Application of machine learning using simple clinical variables and health examination items was helpful to predict incident AF in the general population. Prospective study is warranted to construct an individualized precision medicine.

Download Full-text

Early Prediction of the Carbapenem Resistance Gram-negative Bacteria Carriage in Intensive Care Unit using Machine-Learning

10.21203/rs.3.rs-60222/v1 ◽

2020 ◽

Author(s):

Qiqiang Liang ◽

Qinyu Zhao ◽

Xin Xu ◽

Yu Zhou ◽

Man Huang

Keyword(s):

Machine Learning ◽

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model ◽

Carbapenem Resistance ◽

Multivariate Logistic Regression Model ◽

Gram Negative Bacteria ◽

Multivariate Logistic Regression ◽

Gram Negative ◽

Better Than

Abstract Background The prevention and control of carbapenem-resistance gram-negative bacteria (CR-GNB) is the difficulty and focus for clinicians in the intensive care unit (ICU). This study construct a CR-GNB carriage prediction model in order to predict the CR-GNB incidence in one week. Methods The database is comprised of nearly 10,000 patients. the model is constructed by the multivariate logistic regression model and three machine learning algorithms. Then we choose the optimal model and verify the accuracy by daily predicted and recorded the occurrence of CR-GNB of all patients admitted for 4 months. Results There are 1385 patients with positive CR-GNB cultures and 1535 negative patients in this study. Forty-five variables have statistical significant differences. We include the 17 variables in the multivariate logistic regression model and build three machine learning models for all variables. In terms of accuracy and the area under the receiver operating characteristic (AUROC) curve, the random forest is better than XGBoost and multivariate logistic regression model, and better than decision tree model (accuracy: 84% >82%>81%>72%), (AUROC: 0.9089 > 0.8947 ≈ 0.8987 > 0.7845). In the 4-month prospective study, 81 cases were predicted to be positive in CR-GNB culture within 7 days, 146 cases were predicted to be negative, 86 cases were positive, and 120 cases were negative, with an overall accuracy of 84% and AUROC of 91.98%. Conclusions Prediction models by machine learning can predict the occurrence of CR-GNB colonization or infection within a week period, and can real-time predict and guide medical staff to identify high-risk groups more accurately.

Download Full-text

Effectiveness of Mid-Regional Pro-Adrenomedullin (MR-proADM) as Prognostic Marker in COVID-19 Critically Ill Patients: an Observational Prospective Study

10.21203/rs.3.rs-56715/v1 ◽

2020 ◽

Cited By ~ 1

Author(s):

Giorgia Montrucchio ◽

Gabriele Sales ◽

Francesca Rumbolo ◽

Filippo Palmesino ◽

Vito Fanelli ◽

...

Keyword(s):

Cardiovascular Disease ◽

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model ◽

Predictive Ability ◽

Rank Test ◽

Operating Characteristics ◽

Icu Patients ◽

D Dimer ◽

Over Time

Abstract Background Due to the lack of validated biomarkers to predict disease progression and mortality in COVID-19 ICU-patients, we tested the effectiveness of mid-regional pro-adrenomedullin (MR-proADM) in comparison to C-reactive protein (CRP), procalcitonin (PCT), D-dimer, lactate dehydrogenase (LDH) in predicting outcome.Methods All consecutive COVID-19 adult patients admitted between March and June 2020 to the ICU of the ‘Città della Salute e della Scienza’ hospital in Turin (Italy) were enrolled. MR-proADM, clinical and routine laboratory test were measured within 48 hours from ICU admission, on day 3, 7 and 14. Survival curves difference with MR-proADM cut-off set to 1.8 nmol/L were tested using log-rank test. Predictive ability was compared using area under the curve and 95% confidence interval of different receiver-operating characteristics curves. Potential confounding effects were tested using a logistic regression model. Results Fifty-seven patients were enrolled. ICU and overall mortality were 54.4%. Within the first 24 hours, lymphocytopenia was present in 86%; increased D-dimer and CRP levels were found in 84.2% and 87.7% respectively, while PCT values higher than 0.5 μg/L were observed in 47.4%. MR-proADM, CRP and LDH were significantly different between surviving and non-surviving patients and over time, while PCT, D-dimer and NT-pro-BNP did not show any difference between the groups and over time; lymphocytes count was different between surviving and non-surviving patients only.MR-proADM was higher in dying patients (2.65+2.33vs1.18+0.47, p=0.0001) and a higher mortality characterized patients with MR-proADM exceeding 1.8 nmol/L (p=0.0157). The logistic regression model adjusted for age, gender, cardiovascular disease, diabetes mellitus and PCT values confirmed an odds ratio equal to 10.274 (95%CI 1.970-53.578) (p=0.0057) for MR-proADM higher than 1.8 nmol/L and equal to 22.206 (95%CI 1.56-316.960) (p=0.0223) for cardiovascular disease. Overall, MR-proADM was found to have the best predictive ability (AUC=0.846 – 95%CI 0.779-0.899).Conclusions In COVID-19 ICU-patients, MR-proADM seems able to provide a more precise stratification of disease severity and mortality risk than other biomarkers. Repeated MR-proADM measurement may support a rapid and effective decision-making. Further studies are needed to better explain the mechanisms responsible of the increase in MR-proADM observed in COVID-19 patients.

Download Full-text

Root for a Phishing Page using Machine Learning

International Journal of Innovative Technology and Exploring Engineering - Special Issue ◽

10.35940/ijitee.a5169.119119 ◽

2019 ◽

Vol 9 (1) ◽

pp. 1692-1695

Keyword(s):

Machine Learning ◽

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model

Phishing alludes of the mimicking of the first website. To infiltrate this sort of con,the correspondence claims will a chance to be starting with an official illustrative of a website alternately another institutional Furthermore starting with the place an individual need a probable benefits of the business with. (eg. PayPal,Amazon,UPS,Bank for america etc). It focuses those vunariblities Toward method for pop ups ,ads,fake login pages and so on. Web clients are pulled in Eventually Tom's perusing method for leveraging their trust on acquire their delicate data for example, such that usernames,passwords,account numbers or other data with open record on acquire loans or purchase all the merchandise through e-commerce locales. Upto 5% for clients appear on make lured under these attacks,so it might remain calm gainful for scammers-many about whom who send a large number for trick e-mails An day. In this system,we offer an answer with this issue Toward settling on those client mindful of such phishing exercises Eventually Tom's perusing identifying the trick joins Furthermore urls Toward utilizing the blending of the The majority powerful calculations for machine learning, Concerning illustration An result, we infer our paper with correctness from claiming 98.8% What's more mix from claiming 26 offers. The best algorithm being ,the logistic regression model.

Download Full-text