RESUMO
BACKGROUND: Smoking is a critical risk factor responsible for over eight million annual deaths worldwide. It is essential to obtain information on smoking habits to advance research and implement preventive measures such as screening of high-risk individuals. In most countries, including Denmark, smoking habits are not systematically recorded and at best documented within unstructured free-text segments of electronic health records (EHRs). This would require researchers and clinicians to manually navigate through extensive amounts of unstructured data, which is one of the main reasons that smoking habits are rarely integrated into larger studies. Our aim is to develop machine learning models to classify patients' smoking status from their EHRs. METHODS: This study proposes an efficient natural language processing (NLP) pipeline capable of classifying patients' smoking status and providing explanations for the decisions. The proposed NLP pipeline comprises four distinct components, which are; (1) considering preprocessing techniques to address abbreviations, punctuation, and other textual irregularities, (2) four cutting-edge feature extraction techniques, i.e. Embedding, BERT, Word2Vec, and Count Vectorizer, employed to extract the optimal features, (3) utilization of a Stacking-based Ensemble (SE) model and a Convolutional Long Short-Term Memory Neural Network (CNN-LSTM) for the identification of smoking status, and (4) application of a local interpretable model-agnostic explanation to explain the decisions rendered by the detection models. The EHRs of 23,132 patients with suspected lung cancer were collected from the Region of Southern Denmark during the period 1/1/2009-31/12/2018. A medical professional annotated the data into 'Smoker' and 'Non-Smoker' with further classifications as 'Active-Smoker', 'Former-Smoker', and 'Never-Smoker'. Subsequently, the annotated dataset was used for the development of binary and multiclass classification models. An extensive comparison was conducted of the detection performance across various model architectures. RESULTS: The results of experimental validation confirm the consistency among the models. However, for binary classification, BERT method with CNN-LSTM architecture outperformed other models by achieving precision, recall, and F1-scores between 97% and 99% for both Never-Smokers and Active-Smokers. In multiclass classification, the Embedding technique with CNN-LSTM architecture yielded the most favorable results in class-specific evaluations, with equal performance measures of 97% for Never-Smoker and measures in the range of 86 to 89% for Active-Smoker and 91-92% for Never-Smoker. CONCLUSION: Our proposed NLP pipeline achieved a high level of classification performance. In addition, we presented the explanation of the decision made by the best performing detection model. Future work will expand the model's capabilities to analyze longer notes and a broader range of categories to maximize its utility in further research and screening applications.
Assuntos
Registros Eletrônicos de Saúde , Processamento de Linguagem Natural , Fumar , Humanos , Dinamarca/epidemiologia , Registros Eletrônicos de Saúde/estatística & dados numéricos , Fumar/epidemiologia , Aprendizado de Máquina , Feminino , Masculino , Pessoa de Meia-Idade , Redes Neurais de ComputaçãoRESUMO
The accurate estimation of energy expenditure from simple objective accelerometry measurements provides a valuable method for investigating the effect of physical activity (PA) interventions or population surveillance. Methods have been evaluated previously, but none utilize the temporal aspects of the accelerometry data. In this study, we investigated the energy expenditure prediction from acceleration measured at the subjects' hip, wrist, thigh, and back using recurrent neural networks utilizing temporal elements of the data. The acceleration was measured in children (N = 33) performing a standardized activity protocol in their natural environment. The energy expenditure was modelled using Multiple Linear Regression (MLR), stacked long short-term memory (LSTM) networks, and combined convolutional neural networks (CNN) and LSTM. The correlation and mean absolute percentage error (MAPE) were 0.76 and 19.9% for the MLR, 0.882 and 0.879 and 14.22% for the LSTM, and, with the combined LSTM-CNN, the best performance of 0.883 and 13.9% was achieved. The prediction error for vigorous intensities was significantly different (p < 0.01) from those of the other intensity domains: sedentary, light, and moderate. Utilizing the temporal elements of movement significantly improves energy expenditure prediction accuracy compared to other conventional approaches, but the prediction error for vigorous intensities requires further investigation.
Assuntos
Acelerometria , Metabolismo Energético , Redes Neurais de Computação , Humanos , Acelerometria/métodos , Metabolismo Energético/fisiologia , Masculino , Feminino , Criança , Exercício Físico/fisiologia , Modelos Lineares , Memória de Curto Prazo/fisiologiaRESUMO
In this study, we utilized data from the Surveillance, Epidemiology, and End Results (SEER) database to predict the glioblastoma patients' survival outcomes. To assess dataset skewness and detect feature importance, we applied Pearson's second coefficient test of skewness and the Ordinary Least Squares method, respectively. Using two sampling strategies, holdout and five-fold cross-validation, we developed five machine learning (ML) models alongside a feed-forward deep neural network (DNN) for the multiclass classification and regression prediction of glioblastoma patient survival. After balancing the classification and regression datasets, we obtained 46,340 and 28,573 samples, respectively. Shapley additive explanations (SHAP) were then used to explain the decision-making process of the best model. In both classification and regression tasks, as well as across holdout and cross-validation sampling strategies, the DNN consistently outperformed the ML models. Notably, the accuracy were 90.25% and 90.22% for holdout and five-fold cross-validation, respectively, while the corresponding R2 values were 0.6565 and 0.6622. SHAP analysis revealed the importance of age at diagnosis as the most influential feature in the DNN's survival predictions. These findings suggest that the DNN holds promise as a practical auxiliary tool for clinicians, aiding them in optimal decision-making concerning the treatment and care trajectories for glioblastoma patients.
Assuntos
Aprendizado Profundo , Glioblastoma , Humanos , Glioblastoma/diagnóstico , Bases de Dados Factuais , Hidrolases , Aprendizado de MáquinaRESUMO
AIMS: Modern clinical management of patients with an implantable cardioverter defibrillator (ICD) largely consists of remote device monitoring, although a subset is at risk of mental health issues post-implantation. We compared a 12-month web-based intervention consisting of goal setting, monitoring of patients' mental health-with a psychological intervention if needed-psychoeducational support from a nurse, and an online patient forum, with usual care on participants' device acceptance 12 months after implantation. METHODS AND RESULTS: This national, multi-site, two-arm, non-blinded, randomized, controlled, superiority trial enrolled 478 first-time ICD recipients from all 6 implantation centres in Denmark. The primary endpoint was patient device acceptance measured by the Florida Patient Acceptance Survey (FPAS; general score range = 0-100, with higher scores indicating higher device acceptance) 12 months after implantation. Secondary endpoints included symptoms of depression and anxiety. The primary endpoint of device acceptance was not different between groups at 12 months [B = -2.67, 95% confidence interval (CI) (-5.62, 0.29), P = 0.08]. Furthermore, the secondary endpoint analyses showed no significant treatment effect on either depressive [B = -0.49, 95% CI (-1.19; 0.21), P = 0.17] or anxiety symptoms [B = -0.39, 95% CI (-0.96; 0.18), P = 0.18]. CONCLUSION: The web-based intervention as supplement to usual care did not improve patient device acceptance nor symptoms of anxiety and depression compared with usual care. This specific web-based intervention thus cannot be recommended as a standardized intervention in ICD patients.
Assuntos
Desfibriladores Implantáveis , Qualidade de Vida , Humanos , Ansiedade/prevenção & controle , Ansiedade/psicologia , Atenção à Saúde , Internet , Qualidade de Vida/psicologia , Inquéritos e QuestionáriosRESUMO
BACKGROUND: Alcohol use disorder (AUD) causes significant morbidity, mortality, and injuries. According to reports, approximately 5% of all registered deaths in Denmark could be due to AUD. The problem is compounded by the late identification of patients with AUD, a situation that can cause enormous problems, from psychological to physical to economic problems. Many individuals suffering from AUD never undergo specialist treatment during their addiction due to obstacles such as taboo and the poor performance of current screening tools. Therefore, there is a lack of rapid intervention. This can be mitigated by the early detection of patients with AUD. A clinical decision support system (DSS) powered by machine learning (ML) methods can be used to diagnose patients' AUD status earlier. METHODS: This study proposes an effective AUD prediction model (AUDPM), which can be used in a DSS. The proposed model consists of four distinct components: (1) imputation to address missing values using the k-nearest neighbours approach, (2) recursive feature elimination with cross validation to select the most relevant subset of features, (3) a hybrid synthetic minority oversampling technique-edited nearest neighbour approach to remove noise and balance the distribution of the training data, and (4) an ML model for the early detection of patients with AUD. Two data sources, including a questionnaire and electronic health records of 2571 patients, were collected from Odense University Hospital in the Region of Southern Denmark for the AUD-Dataset. Then, the AUD-Dataset was used to build ML models. The results of different ML models, such as support vector machine, K-nearest neighbour, decision tree, random forest, and extreme gradient boosting, were compared. Finally, a combination of all these models in an ensemble learning approach was selected for the AUDPM. RESULTS: The results revealed that the proposed ensemble AUDPM outperformed other single models and our previous study results, achieving 0.96, 0.94, 0.95, and 0.97 precision, recall, F1-score, and accuracy, respectively. In addition, we designed and developed an AUD-DSS prototype. CONCLUSION: It was shown that our proposed AUDPM achieved high classification performance. In addition, we identified clinical factors related to the early detection of patients with AUD. The designed AUD-DSS is intended to be integrated into the existing Danish health care system to provide novel information to clinical staff if a patient shows signs of harmful alcohol use; in other words, it gives staff a good reason for having a conversation with patients for whom a conversation is relevant.
Assuntos
Alcoolismo , Sistemas de Apoio a Decisões Clínicas , Humanos , Alcoolismo/diagnóstico , Diagnóstico Precoce , Análise por Conglomerados , Registros Eletrônicos de SaúdeRESUMO
BACKGROUND: In May 2022, the World Health Organization (WHO) European Region announced an atypical Monkeypox epidemic in response to reports of numerous cases in some member countries unrelated to those where the illness is endemic. This issue has raised concerns about the widespread nature of this disease around the world. The experience with Coronavirus Disease 2019 (COVID-19) has increased awareness about pandemics among researchers and health authorities. METHODS: Deep Neural Networks (DNNs) have shown promising performance in detecting COVID-19 and predicting its outcomes. As a result, researchers have begun applying similar methods to detect Monkeypox disease. In this study, we utilize a dataset comprising skin images of three diseases: Monkeypox, Chickenpox, Measles, and Normal cases. We develop seven DNN models to identify Monkeypox from these images. Two scenarios of including two classes and four classes are implemented. RESULTS: The results show that our proposed DenseNet201-based architecture has the best performance, with Accuracy = 97.63%, F1-Score = 90.51%, and Area Under Curve (AUC) = 94.27% in two-class scenario; and Accuracy = 95.18%, F1-Score = 89.61%, AUC = 92.06% for four-class scenario. Comparing our study with previous studies with similar scenarios, shows that our proposed model demonstrates superior performance, particularly in terms of the F1-Score metric. For the sake of transparency and explainability, Local Interpretable Model-Agnostic Explanations (LIME) and Gradient-weighted Class Activation Mapping (Grad-Cam) were developed to interpret the results. These techniques aim to provide insights into the decision-making process, thereby increasing the trust of clinicians. CONCLUSION: The DenseNet201 model outperforms the other models in terms of the confusion metrics, regardless of the scenario. One significant accomplishment of this study is the utilization of LIME and Grad-Cam to identify the affected areas and assess their significance in diagnosing diseases based on skin images. By incorporating these techniques, we enhance our understanding of the infected regions and their relevance in distinguishing Monkeypox from other similar diseases. Our proposed model can serve as a valuable auxiliary tool for diagnosing Monkeypox and distinguishing it from other related conditions.
Assuntos
COVID-19 , Mpox , Humanos , COVID-19/diagnóstico , Mpox/diagnóstico , Mpox/epidemiologia , Redes Neurais de Computação , PandemiasRESUMO
Walking ability of elderly individuals, who suffer from walking difficulties, is limited, which restricts their mobility independence. The physical health and well-being of the elderly population are affected by their level of physical activity. Therefore, monitoring daily activities can help improve the quality of life. This becomes especially a huge challenge for those, who suffer from dementia and Alzheimer's disease. Thus, it is of great importance for personnel in care homes/rehabilitation centers to monitor their daily activities and progress. Unlike normal subjects, it is required to place the sensor on the back of this group of patients, which makes it even more challenging to detect walking from other activities. With the latest advancements in the field of health sensing and sensor technology, a huge amount of accelerometer data can be easily collected. In this study, a Machine Learning (ML) based algorithm was developed to analyze the accelerometer data collected from patients with walking difficulties, who live in one of the municipalities in Denmark. The ML algorithm is capable of accurately classifying the walking activity of these individuals with different walking abnormalities. Various statistical, temporal, and spectral features were extracted from the time series data collected using an accelerometer sensor placed on the back of the participants. The back sensor placement is desirable in patients with dementia and Alzheimer's disease since they may remove visible sensors to them due to the nature of their diseases. Then, an evolutionary optimization algorithm called Particle Swarm Optimization (PSO) was used to select a subset of features to be used in the classification step. Four different ML classifiers such as k-Nearest Neighbors (kNN), Random Forest (RF), Stacking Classifier (Stack), and Extreme Gradient Boosting (XGB) were trained and compared on an accelerometry dataset consisting of 20 participants. These models were evaluated using the leave-one-group-out cross-validation (LOGO-CV) technique. The Stack model achieved the best performance with average sensitivity, positive predictive values (precision), F1-score, and accuracy of 86.85%, 93.25%, 88.81%, and 93.32%, respectively, to classify walking episodes. In general, the empirical results confirmed that the proposed models are capable of classifying the walking episodes despite the challenging sensor placement on the back of the patients, who suffer from walking disabilities.
Assuntos
Doença de Alzheimer , Humanos , Idoso , Qualidade de Vida , Caminhada , Marcha , Aprendizado de MáquinaRESUMO
Background: Lung cancer (LC) is the leading cause of cancer related deaths, and several countries are implementing screening programs. Risk models have been introduced to refine the LC screening criteria, but the use of real-world data for this task demands a robust data infrastructure and quality. In this retrospective cohort study, we aim to address the different relevant risk factors in terms of data sources, descriptive statistics, completeness and quality. Methods: Data on comorbidity, prescription medication, smoking history, consultations, symptoms, familial predispositions, exposures, laboratory data among others were collected for all patients examined on a risk of LC over a 10-year period in the Region of Southern Denmark. Data were delivered from the regional data warehouse as well as the Danish Lung Cancer Registry. Associations between LC and non-LC groups were examined through Chi-squared test (categorical variables) and Wilcoxon signed-rank test (continuous variables that were non-parametric). These associations were investigated on both the original datasets and the subset of patients with complete data. Results: The number of examined individuals increased over the study period and more patients were diagnosed with LC in stage I-II, from 18% in 2009 to 31% in 2018. LC patients were more likely to be older, smoker, with a registered prescription of the included medication. They also exhibited differences in laboratory analysis indicating inflammation and hyponatremia. Weight loss, fatigue and pain were more prevalent in the LC group, while hemoptysis and fever were more common among the non-LC patients. Advanced-stage LC patients experienced a higher rate of symptoms compared to those in the low stages. Within the sub-cohort with complete dataset results, most observed trends persisted, although data on comorbidities were susceptibility to change. Conclusions: This study provides key insights into LC risk assessment using a robust dataset of patients examined for suspected LC. A consistent positive trend in early-stage LC diagnosis was observed throughout the study period. LC patients exhibited distinct smoking behaviors, medication patterns, variations in lab results, and specific symptoms. These discoveries have the potential to enhance discrimination in machine learning-based prediction models, particularly those capable of handling complex distributions. Serving as a detailed account of real-world data collection and processing, the study establishes a foundation for future development of prediction models aimed at facilitating the early referral of LC patients.
RESUMO
BACKGROUND: Ovarian cancer is the fifth leading cause of mortality among women in the United States. Ovarian cancer is also known as forgotten cancer or silent disease. The survival of ovarian cancer patients depends on several factors, including the treatment process and the prognosis. METHODS: The ovarian cancer patients' dataset is compiled from the Surveillance, Epidemiology, and End Results (SEER) database. With the help of a clinician, the dataset is curated, and the most relevant features are selected. Pearson's second coefficient of skewness test is used to evaluate the skewness of the dataset. Pearson correlation coefficient is also used to investigate the associations between features. Statistical test is utilized to evaluate the significance of the features. Six Machine Learning (ML) models, including K-Nearest Neighbors , Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), Adaptive Boosting (AdaBoost), and Extreme Gradient Boosting (XGBoost), are implemented for survival prediction in both classification and regression approaches. An interpretable method, Shapley Additive Explanations (SHAP), is applied to clarify the decision-making process and determine the importance of each feature in prediction. Additionally, DTs of the RF model are displayed to show how the model predicts the survival intervals. RESULTS: Our results show that RF (Accuracy = 88.72%, AUC = 82.38%) and XGBoost (Root Mean Squad Error (RMSE)) = 20.61%, R2 = 0.4667) have the best performance for classification and regression approaches, respectively. Furthermore, using the SHAP method along with extracted DTs of the RF model, the most important features in the dataset are identified. Histologic type ICD-O-3, chemotherapy recode, year of diagnosis, age at diagnosis, tumor stage, and grade are the most important determinant factors in survival prediction. CONCLUSION: To the best of our knowledge, our study is the first study that develops various ML models to predict ovarian cancer patients' survival on the SEER database in both classification and regression approaches. These ML algorithms also achieve more accurate results and outperform statistical methods. Furthermore, our study is the first study to use the SHAP method to increase confidence and transparency of the proposed models' prediction for clinicians. Moreover, our developed models, as an automated auxiliary tool, can help clinicians to have a better understanding of the estimated survival as well as important features that affect survival.
Assuntos
Aprendizado de Máquina , Neoplasias Ovarianas , Humanos , Feminino , Neoplasias Ovarianas/diagnóstico , Algoritmos , Prognóstico , Algoritmo Florestas AleatóriasRESUMO
BACKGROUND: High dimensionality in electronic health records (EHR) causes a significant computational problem for any systematic search for predictive, diagnostic, or prognostic patterns. Feature selection (FS) methods have been indicated to be effective in feature reduction as well as in identifying risk factors related to prediction of clinical disorders. This paper examines the prediction of patients with alcohol use disorder (AUD) using machine learning (ML) and attempts to identify risk factors related to the diagnosis of AUD. METHODS: A FS framework consisting of two operational levels, base selectors and ensemble selectors. The first level consists of five FS methods: three filter methods, one wrapper method, and one embedded method. Base selector outputs are aggregated to develop four ensemble FS methods. The outputs of FS method were then fed into three ML algorithms: support vector machine (SVM), K-nearest neighbor (KNN), and random forest (RF) to compare and identify the best feature subset for the prediction of AUD from EHRs. RESULTS: In terms of feature reduction, the embedded FS method could significantly reduce the number of features from 361 to 131. In terms of classification performance, RF based on 272 features selected by our proposed ensemble method (Union FS) with the highest accuracy in predicting patients with AUD, 96%, outperformed all other models in terms of AUROC, AUPRC, Precision, Recall, and F1-Score. Considering the limitations of embedded and wrapper methods, the best overall performance was achieved by our proposed Union Filter FS, which reduced the number of features to 223 and improved Precision, Recall, and F1-Score in RF from 0.77, 0.65, and 0.71 to 0.87, 0.81, and 0.84, respectively. Our findings indicate that, besides gender, age, and length of stay at the hospital, diagnosis related to digestive organs, bones, muscles and connective tissue, and the nervous systems are important clinical factors related to the prediction of patients with AUD. CONCLUSION: Our proposed FS method could improve the classification performance significantly. It could identify clinical factors related to prediction of AUD from EHRs, thereby effectively helping clinical staff to identify and treat AUD patients and improving medical knowledge of the AUD condition. Moreover, the diversity of features among female and male patients as well as gender disparity were investigated using FS methods and ML techniques.
Assuntos
Alcoolismo , Humanos , Masculino , Feminino , Alcoolismo/diagnóstico , Registros Eletrônicos de Saúde , Aprendizado de Máquina , Análise por Conglomerados , Máquina de Vetores de SuporteRESUMO
Introduction: Emergency departments (ED) at hospitals sometimes experience unexpected deterioration in patients that were assessed to be in a stable condition upon arrival. Odense University Hospital (OUH) has conducted a retrospective study to investigate the possibilities of prognostic tools that can detect these unexpected deterioration cases at an earlier stage. The study suggests that the temperature difference (gradient) between the core and the peripheral body parts can be used to detect these cases. The temperature between the patient's inner canthus (core temperature) and the tip of the nose (peripheral temperature) can be measured with a thermal camera. Based on the temperature measurement from a thermal image, a gradient value can be calculated, which can be used as an early indicator of potential deterioration. Problem: The lack of a tool to automatically calculate the gradient has prevented the ED at OUH in conducting a comprehensive prospective study on early indicators of patients at risk of deterioration. The current manual way of doing facial landmark detection on thermal images is too time consuming and not feasible as part of the daily workflow at the ED, where nurses have to triage patients within a few minutes. Objective: The objective of this study was to automate the process of calculating the gradient by developing a handheld prognostic tool that can be used by nurses for automatically performing facial landmark detection on thermal images of patients as they arrive at the ED. Methods: A systematic literature review has been conducted to investigate previous studies that have been done for applying computer vision methods on thermal images. Several meetings, interviews and field studies have been conducted with the ED at OUH in order to understand their workflow, formulate and prioritize requirements and co-design the prognostic tool. Results: The study resulted in a novel Android app that can capture a thermal image of a patient's face with a thermal camera attached to a smartphone. Within a few seconds, the app then automatically calculates the gradient to be used in the triage process. The developed tool is the first of its kind using facial landmark detection on thermal images for calculating a gradient that can serve as a novel prognostic indicator for ED patients.
RESUMO
BACKGROUND: Atrial fibrillation (AF) is one of the most prevalent cardiac arrhythmias, which challenges the healthcare systems globally.Timely detection of AF can potentially reduce the mortality and morbidity rates as well as alleviate the economic burden caused by this.Digital solutions are shown to enhance the diagnosis of cardiac abnormalities. OBJECTIVES: By the latest advancements in the field of medical informatics and tele-health monitoring, huge amount of electro-physiological signals, such as electrocardiograms (ECG), can be easily collected.One of the most common ways for physicians/cardiologists to analyse these signals is through visual inspection.However, it is not always easy and in most cases cumbersome to analyse these big amounts of ECG data.Therefore, it is of great interest to develop models that are capable of analyzing these data and help physicians making better decisions.This paper proposes and compares well-known machine learning (ML) algorithms to diagnose short episodes of AF. This also paves the way for real-time detection of AF in clinical settings. METHODS: Different ML algorithms such as Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), Stacking Classifier (SC), Extreme Gradient Boosting (XGBoost), and Adaptive Boosting (AdaBoost) were applied to detect AF. These models were trained using extracted statistical features from ECG signals. RESULTS: The proposed ML models were trained on a dataset with 23 ECG records of length approximately 10 h each using leave one group out cross validation (LOGO-CV) technique and achieved the best sensitivity (Se), specificity (Sp), positive predictive value (PPV), false positive rate (FPR), and F1-score of 85.67%, 81.25%, 90.85%, 18.75% and 88.18%, respectively, to classify AF from normal sinus rhythms (NSR) in short ECG segments of 20 heartbeats.Additionally, the models were examined on three unseen datasets, namely the Long Term AF dataset, MIT-BIH Arrhythmia dataset, and MIT-BIH Normal Sinus Rhythm dataset, to assess their robustness and generalization. CONCLUSION: The obtained results show high performance and flexibility of some of the applied ML models compared to other well-known algorithms. In general, the empirical results confirm that ensemble methods, such as AdaBoost, generalized well and perform better than other approaches.
Assuntos
Fibrilação Atrial , Algoritmos , Fibrilação Atrial/diagnóstico , Eletrocardiografia/métodos , Frequência Cardíaca , Humanos , Máquina de Vetores de SuporteRESUMO
OBJECTIVES: This systematic review aimed to assess the performance and clinical feasibility of machine learning (ML) algorithms in prediction of in-hospital mortality for medical patients using vital signs at emergency departments (EDs). DESIGN: A systematic review was performed. SETTING: The databases including Medline (PubMed), Scopus and Embase (Ovid) were searched between 2010 and 2021, to extract published articles in English, describing ML-based models utilising vital sign variables to predict in-hospital mortality for patients admitted at EDs. Critical appraisal and data extraction for systematic reviews of prediction modelling studies checklist was used for study planning and data extraction. The risk of bias for included papers was assessed using the prediction risk of bias assessment tool. PARTICIPANTS: Admitted patients to the ED. MAIN OUTCOME MEASURE: In-hospital mortality. RESULTS: Fifteen articles were included in the final review. We found that eight models including logistic regression, decision tree, K-nearest neighbours, support vector machine, gradient boosting, random forest, artificial neural networks and deep neural networks have been applied in this domain. Most studies failed to report essential main analysis steps such as data preprocessing and handling missing values. Fourteen included studies had a high risk of bias in the statistical analysis part, which could lead to poor performance in practice. Although the main aim of all studies was developing a predictive model for mortality, nine articles did not provide a time horizon for the prediction. CONCLUSION: This review provided an updated overview of the state-of-the-art and revealed research gaps; based on these, we provide eight recommendations for future studies to make the use of ML more feasible in practice. By following these recommendations, we expect to see more robust ML models applied in the future to help clinicians identify patient deterioration earlier.
Assuntos
Serviço Hospitalar de Emergência , Aprendizado de Máquina , Hospitalização , Humanos , Modelos LogísticosRESUMO
BACKGROUND: Prediction of length of stay (LOS) at admission time can provide physicians and nurses insight into the illness severity of patients and aid them in avoiding adverse events and clinical deterioration. It also assists hospitals with more effectively managing their resources and manpower. METHODS: In this field of research, there are some important challenges, such as missing values and LOS data skewness. Moreover, various studies use a binary classification which puts a wide range of patients with different conditions into one category. To address these shortcomings, first multivariate imputation techniques are applied to fill incomplete records, then two proper resampling techniques, namely Borderline-SMOTE and SMOGN, are applied to address data skewness in the classification and regression domains, respectively. Finally, machine learning (ML) techniques including neural networks, extreme gradient boosting, random forest, support vector machine, and decision tree are implemented for both approaches to predict LOS of patients admitted to the Emergency Department of Odense University Hospital between June 2018 and April 2019. The ML models are developed based on data obtained from patients at admission time, including pulse rate, arterial blood oxygen saturation, respiratory rate, systolic blood pressure, triage category, arrival ICD-10 codes, age, and gender. RESULTS: The performance of predictive models before and after addressing missing values and data skewness is evaluated using four evaluation metrics namely receiver operating characteristic, area under the curve (AUC), R-squared score (R2), and normalized root mean square error (NRMSE). Results show that the performance of predictive models is improved on average by 15.75% for AUC, 32.19% for R2 score, and 11.32% for NRMSE after addressing the mentioned challenges. Moreover, our results indicate that there is a relationship between the missing values rate, data skewness, and illness severity of patients, so it is clinically essential to take incomplete records of patients into account and apply proper solutions for interpolation of missing values. CONCLUSION: We propose a new method comprised of three stages: missing values imputation, data skewness handling, and building predictive models based on classification and regression approaches. Our results indicated that addressing these challenges in a proper way enhanced the performance of models significantly, which led to a more valid prediction of LOS.
Assuntos
Serviço Hospitalar de Emergência , Triagem , Hospitalização , Humanos , Tempo de Internação , Aprendizado de MáquinaRESUMO
BACKGROUND: As a preamble to an attempt to develop a tool that can aid health professionals at hospitals in identifying whether the patient may have an alcohol abuse problem, this study investigates opinions and attitudes among both health professionals and patients about using patient data from electronic health records (EHRs) in an algorithm screening for alcohol problems. OBJECTIVE: The aim of this study was to investigate the attitudes and opinions of patients and health professionals at hospitals regarding the use of previously collected data in developing and implementing an algorithmic helping tool in EHR for screening inexpedient alcohol habits; in addition, the study aims to analyze how patients would feel about asking and being asked about alcohol by staff, based on a notification in the EHR from such a tool. METHODS: Using semistructured interviews, we interviewed 9 health professionals and 5 patients to explore their opinions and attitudes about an algorithm-based helping tool and about asking and being asked about alcohol usage when being given a reminder from this type of tool. The data were analyzed using an ad hoc method consistent with a close reading and meaning condensing. RESULTS: The health professionals were both positive and negative about a helping tool grounded in algorithms. They were optimistic about the potential of such a tool to save some time by providing a quick overview if it was easy to use but, on the negative side, noted that this type of helping tool might take away the professionals' instinct. The patients were overall positive about the helping tool, stating that they would find this tool beneficial for preventive care. Some of the patients expressed concerns that the information provided by the tool could be misused. CONCLUSIONS: When developing and implementing an algorithmic helping tool, the following aspects should be considered: (1) making the helping tool as transparent in its recommendations as possible, avoiding black boxing, and ensuring room for professional discretion in clinical decision making; and (2) including and taking into account the attitudes and opinions of patients and health professionals in the design and development process of such an algorithmic helping tool.
RESUMO
COVID-19 has had a severe impact globally, and the recovery can be characterized as a tug of war between fast economic recovery and firm control of further virus-spread. To be prepared for future pandemics, public health policy makers should put effort into fully understanding any complex psychological tensions that inherently arise between opposing human factors such as free enjoyment versus self-restriction. As the COVID-19 crisis is an unusual and complex problem, combinations of diverse factors such as health risk perception, knowledge, norms and beliefs, attitudes and behaviors are closely associated with individuals' intention to enjoy the experience economy but also their concerns that the experience economy will trigger further spread of the infectious diseases. Our aim is to try identifying what factors are associated with their concerns about the spread of the infectious disease caused by the local experience economy. Hence, we have chosen a "data-driven" explanatory approach, "Probabilistic Structural Equational Modeling," based on the principle of Bayesian networks to analyze data collected from the following four countries with indicated sample sizes: Denmark (1,005), Italy (1,005), China (1,013), and Japan (1,091). Our findings highlight the importance of understanding the contextual differences in relations between the target variable and factors such as personal value priority and knowledge. These factors affect the target variable differently depending on the local severity-level of the infections. Relations between pleasure-seeking via the experience economy and individuals' anxiety-level about an infectious hotspot seem to differ between East Asians and Europeans who are known to prioritize so-called interpersonal- and independent self-schemes, respectively. Our study also indicates the heterogeneity in the populations, i.e., these relations differ within the respective populations. Another finding shows that the Japanese population is particularly concerned about their local community potentially becoming an infectious hotspot and hence expecting others to comply with their particular social norms. Summarizing, the obtained insights imply the importance of considering both cultural- and individual contexts when policy makers are going to develop measures to address pandemic dilemmas such as maintaining public health awareness and accelerating the recovery of the local experience economy.
RESUMO
The volume of daily patient arrivals at Emergency Departments (EDs) is unpredictable and is a significant reason of ED crowding in hospitals worldwide. Timely forecast of patients arriving at ED can help the hospital management in early planning and avoiding of overcrowding. Many different ED patient arrivals forecasting models have previously been proposed by using time series analysis methods. Even though the time series methods such as Linear and Logistic Regression, Autoregressive Integrated Moving Average (ARIMA), Seasonal ARIMA (SARIMA), Exponential Smoothing (ES), and Artificial Neural Network (ANN) have been explored extensively for the ED forecasting model development, the few significant limitations of these methods associated in the analysis of time series data make the models inadequate in many practical situations. Therefore, in this paper, Machine Learning (ML)-based Random Forest (RF) regressor, and Deep Neural Network (DNN)-based Long Short-Term Memory (LSTM) and Convolutional Neural network (CNN) methods, which have not been explored to the same extent as the other time series techniques, are implemented by incorporating meteorological and calendar parameters for the development of forecasting models. The performances of the developed three models in forecasting ED patient arrivals are evaluated. Among the three models, CNN outperformed for short-term (3 days in advance) patient arrivals prediction with Mean Absolute Percentage Error (MAPE) of 9.24% and LSTM performed better for moderate-term (7 days in advance) patient arrivals prediction with MAPE of 8.91% using weather forecast information. Whereas, LSTM model outperformed with MAPE of 8.04% compared to 9.53% by CNN and 10.10% by RF model for current day prediction of patient arrivals using 3 days past weather information. Thus, for short-term ED patient arrivals forecasting, DNN-based model performed better compared to RF regressor ML-based model.
Assuntos
Serviço Hospitalar de Emergência , Redes Neurais de Computação , Previsões , Humanos , Aprendizado de MáquinaRESUMO
This paper presents an application of deep neural networks (DNN) to identify patients with Alcohol Use Disorder based on historical electronic health records. Our methodology consists of four stages including data collection, preprocessing, predictive model development, and validation. Data are collected from two sources and labeled into three classes including Normal, Hazardous, and Harmful drinkers. Moreover, problems such as imbalanced classes, noise, and categorical variables were handled. A four-layer fully-connected feedforward DNN architecture was designed and developed to predict Normal, Hazardous, and Harmful drinkers. Results show that our proposed method could successfully classify about 96%, 82%, and 89% of Normal, Hazardous, and Harmful drinkers, respectively, which is better than classical machine learning approaches.