A machine learning-based framework to identify type 2 diabetes through electronic health records

Applying machine learning approaches for predicting obesity risk using US health administrative claims database

INTRODUCTION

Body mass index (BMI) is inadequately recorded in US administrative claims databases. We aimed to validate the sensitivity and positive predictive value (PPV) of BMI-related diagnosis codes using an electronic medical records (EMR) claims-linked database. Additionally, we applied machine learning (ML) to identify features in US claims databases to predict obesity status.

RESEARCH DESIGN AND METHODS

This observational, retrospective analysis included 692 119 people ≥18 years of age, with ≥1 BMI reading in MarketScan Explorys Claims-EMR data (January 2013-December 2019). Claims-based obesity status was compared with EMR-based BMI (gold standard) to assess BMI-related diagnosis code sensitivity and PPV. Logistic regression (LR), penalized LR with L1 penalty (Least Absolute Shrinkage and Selection Operator), extreme gradient boosting (XGBoost) and random forest, with features drawn from insurance claims, were trained to predict obesity status (BMI≥30 kg/m2) from EMR as the gold standard. Model performance was compared using several metrics, including the area under the receiver operating characteristic curve. The best-performing model was applied to assess feature importance. Obesity risk scores were computed from the best model generated from the claims database and compared against the BMI recorded in the EMR.

RESULTS

The PPV of diagnosis codes from claims alone remained high over the study period (85.4-89.2%); sensitivity was low (16.8-44.8%). XGBoost performed the best at predicting obesity with the highest area under the curve (AUC; 79.4%) and the lowest Brier score. The number of obesity diagnoses and obesity diagnoses from inpatient settings were the most important predictors of obesity. XGBoost showed an AUC of 74.1% when trained without an obesity diagnosis.

CONCLUSIONS

Obesity prevalence is under-reported in claims databases. ML models, with or without explicit obesity, show promise in improving obesity prediction accuracy compared with obesity codes alone. Improved obesity status prediction may assist practitioners and payors to estimate the burden of obesity and investigate the potential unmet needs of current treatments.

Utilizing machine learning to classify persistent organic pollutants in the serum of pregnant women: a predictive modeling approach

Polychlorinated biphenyls (PCBs), organochlorine pesticides (OCPs), polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/Fs), and per- and poly-fluoroalkyl substances (PFAS) are persistent organic pollutants (POPs) that remain detrimental to critical subpopulations, namely pregnant women. Required tests for biomonitoring are quite expensive. Moreover, statistical models aiming to discover the relationships between pollutants levels and human characteristics have their limitations. Therefore, the objective of this study is to use machine learning predictive models to further examine the pollutants' predictors, while comparing them. Levels of 33 congeners were measured in the serum of 269 pregnant women, from whom data was collected regarding sociodemographic, dietary, environmental, and anthropometric characteristics. Several machine learning algorithms were compared using "Python" for each pollutant: support vector machine (SVM), random forest, XGBoost, and neural networks. The aforementioned characteristics were included in the model as features. Prediction, accuracy, precision, recall, F1-score, area under the ROC curve (AUC), sensitivity, and specificity were retrieved to compare the models between them and among pollutants. The highest performing model for all pollutants was Random Forest. Results showed a moderate to acceptable performance and discriminative power among all POPs, with OCPs' model performing slightly better than all other models. Top related features for each model were also presented using SHAP analysis, detailing the predictors' negative or positive impact on the model. In conclusion, developing such a tool is of major importance in a context of limited financial and research resources. Nevertheless, machine learning models should always be interpreted with caution by exploring all evaluation metrics.

Graph neural networks for clinical risk prediction based on electronic health records: A survey

OBJECTIVE

This study aims to comprehensively review the use of graph neural networks (GNNs) for clinical risk prediction based on electronic health records (EHRs). The primary goal is to provide an overview of the state-of-the-art of this subject, highlighting ongoing research efforts and identifying existing challenges in developing effective GNNs for improved prediction of clinical risks.

METHODS

A search was conducted in the Scopus, PubMed, ACM Digital Library, and Embase databases to identify relevant English-language papers that used GNNs for clinical risk prediction based on EHR data. The study includes original research papers published between January 2009 and May 2023.

RESULTS

Following the initial screening process, 50 articles were included in the data collection. A significant increase in publications from 2020 was observed, with most selected papers focusing on diagnosis prediction (n = 36). The study revealed that the graph attention network (GAT) (n = 19) was the most prevalent architecture, and MIMIC-III (n = 23) was the most common data resource.

CONCLUSION

GNNs are relevant tools for predicting clinical risk by accounting for the relational aspects among medical events and entities and managing large volumes of EHR data. Future studies in this area may address challenges such as EHR data heterogeneity, multimodality, and model interpretability, aiming to develop more holistic GNN models that can produce more accurate predictions, be effectively implemented in clinical settings, and ultimately improve patient care.

Machine learning to identify chronic cough from administrative claims data

Accurate identification of patient populations is an essential component of clinical research, especially for medical conditions such as chronic cough that are inconsistently defined and diagnosed. We aimed to develop and compare machine learning models to identify chronic cough from medical and pharmacy claims data. In this retrospective observational study, we compared 3 machine learning algorithms based on XG Boost, logistic regression, and neural network approaches using a large claims and electronic health record database. Of the 327,423 patients who met the study criteria, 4,818 had chronic cough based on linked claims-electronic health record data. The XG Boost model showed the best performance, achieving a Receiver-Operator Characteristic Area Under the Curve (ROC-AUC) of 0.916. We selected a cutoff that favors a high positive predictive value (PPV) to minimize false positives, resulting in a sensitivity, specificity, PPV, and negative predictive value of 18.0%, 99.6%, 38.7%, and 98.8%, respectively on the held-out testing set (n = 82,262). Logistic regression and neural network models achieved slightly lower ROC-AUCs of 0.907 and 0.838, respectively. The XG Boost and logistic regression models maintained their robust performance in subgroups of individuals with higher rates of chronic cough. Machine learning algorithms are one way of identifying conditions that are not coded in medical records, and can help identify individuals with chronic cough from claims data with a high degree of classification value.

Application of a Machine Learning Algorithm in Prediction of Abusive Head Trauma in Children

PURPOSE

We explored the application of a machine learning algorithm for the timely detection of potential abusive head trauma (AHT) using the first free-text note of an encounter and demographic information.

METHODS

First free-text physician notes and demographic information were collected for children under 5 years of age at a Level 1 Trauma Center. The control group, which included patients with head/neck injury, was compared to those with AHT diagnosed by the Child Protective Team. Differential scores accounted for words overrepresented in AHT patient vs. control notes. Sentiment scores were reflective of note positivity/negativity and subjectivity scores accounted for note subjectivity/objectivity. The composite scores reflected the patient's differential score modified by the subjectivity score. Composite, sentiment, and subjectivity scores combined with demographic information trained a Random Forest (RF) machine learning algorithm to predict AHT.

RESULTS

Final composite scores with demographic information were highly associated with AHT in a test dataset. The control group included 587 patients and the test group included 193 patients. Combining composite scores with demographic information into the RF model improved AHT classification area under the curve (AUC) from 0.68 to 0.78, with an overall accuracy of 84%. Feature importance analysis of our RF model revealed that composite score, sentiment, age, and subjectivity were the most impactful predictors of AHT. The sentiment was not significantly different between control and AHT notes (p = 0.87), while subjectivity trended higher for AHT notes (p = 0.081).

CONCLUSION

We conclude that a machine learning algorithm can recognize patterns within free-text notes and demographic information that aid in AHT detection in children.

LEVEL OF EVIDENCE

III.

Nanomaterials for detection of biomolecules and delivering therapeutic agents in theragnosis: A review

Nanomaterials are emerging facts used to deliver therapeutic agents in living systems. Nanotechnology is used as a compliment by implementing different kinds of nanotechnological applications such as nano-porous structures, functionalized nanomaterials, quantum dots, carbon nanomaterials, and polymeric nanostructures. The applications are in the initial stage, which led to achieving several diagnoses and therapy in clinical practice. This review conveys the importance of nanomaterials in post-genomic employment, which includes the design of immunosensors, immune assays, and drug delivery. In this view, genomics is a molecular tool containing large databases that are useful in choosing an apt molecular inhibitor such as drug, ligand and antibody target in the drug delivery process. This study identifies the expression of genes and proteins in analysis and classification of diseases. Experimentally, the study analyses the design of a disease model. In particular, drug delivery is a boon area to treat cancer. The identified drugs enter different phase trails (Trails I, II, and III). The genomic information conveys more essential entities to the phase I trials and helps to move further for other trails such as trails-II and III. In such cases, the biomarkers play a crucial role by monitoring the unique pathological process. Genetic engineering with recombinant DNA techniques can be employed to develop genetically engineered disease models. Delivering drugs in a specific area is one of the challenging issues achieved using nanoparticles. Therefore, genomics is considered as a vast molecular tool to identify drugs in personalized medicine for cancer therapy.

A Machine Learning Web App to Predict Diabetic Blood Glucose Based on a Basic Noninvasive Health Checkup, Sociodemographic Characteristics, and Dietary Information: Case Study

BACKGROUND

Over the past few decades, diabetes has become a serious public health concern worldwide, particularly in Bangladesh. The advancement of artificial intelligence can be reaped in the prediction of blood glucose levels for better health management. However, the practical validity of machine learning (ML) techniques for predicting health parameters using data from low- and middle-income countries, such as Bangladesh, is very low. Specifically, Bangladesh lacks research using ML techniques to predict blood glucose levels based on basic noninvasive clinical measurements and dietary and sociodemographic information.

OBJECTIVE

To formulate strategies for public health planning and the control of diabetes, this study aimed to develop a personalized ML model that predicts the blood glucose level of urban corporate workers in Bangladesh.

METHODS

Based on the basic noninvasive health checkup test results, dietary information, and sociodemographic characteristics of 271 employees of the Bangladeshi Grameen Bank complex, 5 well-known ML models, namely, linear regression, boosted decision tree regression, neural network, decision forest regression, and Bayesian linear regression, were used to predict blood glucose levels. Continuous blood glucose data were used in this study to train the model, which then used the trained data to predict new blood glucose values.

RESULTS

Boosted decision tree regression demonstrated the greatest predictive performance of all evaluated models (root mean squared error=2.30). This means that, on average, our model's predicted blood glucose level deviated from the actual blood glucose level by around 2.30 mg/dL. The mean blood glucose value of the population studied was 128.02 mg/dL (SD 56.92), indicating a borderline result for the majority of the samples (normal value: 140 mg/dL). This suggests that the individuals should be monitoring their blood glucose levels regularly.

CONCLUSIONS

This ML-enabled web application for blood glucose prediction helps individuals to self-monitor their health condition. The application was developed with communities in remote areas of low- and middle-income countries, such as Bangladesh, in mind. These areas typically lack health facilities and have an insufficient number of qualified doctors and nurses. The web-based application is a simple, practical, and effective solution that can be adopted by the community. Use of the web application can save money on medical expenses, time, and health management expenses. The created system also aids in achieving the Sustainable Development Goals, particularly in ensuring that everyone in the community enjoys good health and well-being and lowering total morbidity and mortality.

Establishment of a differential diagnosis method and an online prediction platform for AOSD and sepsis based on gradient boosting decision trees algorithm

OBJECTIVE

The differential diagnosis between adult-onset Still's disease (AOSD) and sepsis has always been a challenge. In this study, a machine learning model for differential diagnosis of AOSD and sepsis was developed and an online platform was developed to facilitate the clinical application of the model.

METHODS

All data were collected from 42 AOSD patients and 50 sepsis patients admitted to Affiliated Hospital of Xuzhou Medical University from December 2018 to December 2021. In addition, 5 AOSD patients and 10 sepsis patients diagnosed in our hospital after March 2022 were collected for external validation. All models were built using the scikit-learn library (version 1.0.2) in Python (version 3.9.7), and feature selection was performed using the SHAP (Shapley Additive exPlanation) package developed in Python.

RESULTS

The results showed that the gradient boosting decision tree(GBDT) optimization model based on arthralgia, ferritin × lymphocyte count, white blood cell count, ferritin × platelet count, and α1-acid glycoprotein/creatine kinase could well identify AOSD and sepsis. The training set interaction test (AUC: 0.9916, ACC: 0.9457, Sens: 0.9556, Spec: 0.9578) and the external validation also achieved satisfactory results (AUC: 0.9800, ACC: 0.9333, Sens: 0.8000, Spec: 1.000). We named this discrimination method AIADSS (AI-assisted discrimination of Still's disease and Sepsis) and created an online service platform for practical operation, the website is http://cppdd.cn/STILL1/ .

CONCLUSION

We created a method for the identification of AOSD and sepsis based on machine learning. This method can provide a reference for clinicians to formulate the next diagnosis and treatment plan.

Artificial Intelligence (AI) in pediatric endocrinology

Artificial Intelligence (AI) is integrating itself throughout the medical community. AI's ability to analyze complex patterns and interpret large amounts of data will have considerable impact on all areas of medicine, including pediatric endocrinology. In this paper, we review and update the current studies of AI in pediatric endocrinology. Specific topics that are addressed include: diabetes management, bone growth, metabolism, obesity, and puberty. Becoming knowledgeable and comfortable with AI will assist pediatric endocrinologists, the goal of the paper.

Clinical named entity recognition and relation extraction using natural language processing of medical free text: A systematic review

BACKGROUND

Natural Language Processing (NLP) applications have developed over the past years in various fields including its application to clinical free text for named entity recognition and relation extraction. However, there has been rapid developments the last few years that there's currently no overview of it. Moreover, it is unclear how these models and tools have been translated into clinical practice. We aim to synthesize and review these developments.

METHODS

We reviewed literature from 2010 to date, searching PubMed, Scopus, the Association of Computational Linguistics (ACL), and Association of Computer Machinery (ACM) libraries for studies of NLP systems performing general-purpose (i.e., not disease- or treatment-specific) information extraction and relation extraction tasks in unstructured clinical text (e.g., discharge summaries).

RESULTS

We included in the review 94 studies with 30 studies published in the last three years. Machine learning methods were used in 68 studies, rule-based in 5 studies, and both in 22 studies. 63 studies focused on Named Entity Recognition, 13 on Relation Extraction and 18 performed both. The most frequently extracted entities were "problem", "test" and "treatment". 72 studies used public datasets and 22 studies used proprietary datasets alone. Only 14 studies defined clearly a clinical or information task to be addressed by the system and just three studies reported its use outside the experimental setting. Only 7 studies shared a pre-trained model and only 8 an available software tool.

DISCUSSION

Machine learning-based methods have dominated the NLP field on information extraction tasks. More recently, Transformer-based language models are taking the lead and showing the strongest performance. However, these developments are mostly based on a few datasets and generic annotations, with very few real-world use cases. This may raise questions about the generalizability of findings, translation into practice and highlights the need for robust clinical evaluation.

Algorithms to define diabetes type using data from administrative databases: A systematic review of the evidence

AIMS

To find the best-performing algorithms to distinguish type 1 and type 2 diabetes in administrative data.

METHODS

Embase and MEDLINE databases were searched from January 2000 until January 2023. Papers evaluating the performance of algorithms to define type 1 and type 2 diabetes by reporting diagnostic metrics against a range of reference standards were selected. Study quality was evaluated using the Quality Assessment of Diagnostic Accuracy Studies.

RESULTS

Of the 24 studies meeting the eligibility criteria, 19 demonstrated a low risk of bias and low concerns about the applicability of the study population across all domains. Algorithms considering multiple diabetes diagnostic codes alone were sensitive and specific approaches to classify diabetes type (both metrics >92.1% for type 1 diabetes; >86.9% for type 2 diabetes). Among the top 10-performing algorithms to detect type 1 and type 2 diabetes, 70% and 100% featured multiple criteria, respectively. Information on insulin use was more sensitive and specific for detecting diabetes type than were criteria based on use of oral hypoglycaemic agents.

CONCLUSIONS

Algorithms based on multiple diabetes diagnostic codes and insulin use are the most accurate approaches to distinguish type 1 from type 2 diabetes using administrative data. Approaches with more than one criterion may also increase sensitivity in distinguishing diabetes type.

Effect of Artificial Intelligence-based Health Education Accurately Linking System (AI-HEALS) for Type 2 diabetes self-management: protocol for a mixed-methods study

BACKGROUND

Patients with type 2 diabetes (T2DM) have an increasing need for personalized and Precise management as medical technology advances. Artificial intelligence (AI) technologies on mobile devices are being developed gradually in a variety of healthcare fields. As an AI field, knowledge graph (KG) is being developed to extract and store structured knowledge from massive data sets. It has great prospects for T2DM medical information retrieval, clinical decision-making, and individual intelligent question and answering (QA), but has yet to be thoroughly researched in T2DM intervention. Therefore, we designed an artificial intelligence-based health education accurately linking system (AI-HEALS) to evaluate if the AI-HEALS-based intervention could help patients with T2DM improve their self-management abilities and blood glucose control in primary healthcare.

METHODS

This is a nested mixed-method study that includes a community-based cluster-randomized control trial and personal in-depth interviews. Individuals with T2DM between the ages of 18 and 75 will be recruited from 40-45 community health centers in Beijing, China. Participants will either receive standard diabetes primary care (SDPC) (control, 3 months) or SDPC plus AI-HEALS online health education program (intervention, 3 months). The AI-HEALS runs in the WeChat service platform, which includes a KBQA, a system of physiological indicators and lifestyle recording and monitoring, medication and blood glucose monitoring reminders, and automated, personalized message sending. Data on sociodemography, medical examination, blood glucose, and self-management behavior will be collected at baseline, as well as 1,3,6,12, and 18 months later. The primary outcome is to reduce HbA1c levels. Secondary outcomes include changes in self-management behavior, social cognition, psychology, T2DM skills, and health literacy. Furthermore, the cost-effectiveness of the AI-HEALS-based intervention will be evaluated.

DISCUSSION

KBQA system is an innovative and cost-effective technology for health education and promotion for T2DM patients, but it is not yet widely used in the T2DM interventions. This trial will provide evidence on the efficacy of AI and mHealth-based personalized interventions in primary care for improving T2DM outcomes and self-management behaviors.

TRIAL REGISTRATION

Biomedical Ethics Committee of Peking University: IRB00001052-22,058, 2022/06/06; Clinical Trials: ChiCTR2300068952, 02/03/2023.

Phenomic Studies on Diseases: Potential and Challenges

The rapid development of such research field as multi-omics and artificial intelligence (AI) has made it possible to acquire and analyze the multi-dimensional big data of human phenomes. Increasing evidence has indicated that phenomics can provide a revolutionary strategy and approach for discovering new risk factors, diagnostic biomarkers and precision therapies of diseases, which holds profound advantages over conventional approaches for realizing precision medicine: first, the big data of patients' phenomes can provide remarkably richer information than that of the genomes; second, phenomic studies on diseases may expose the correlations among cross-scale and multi-dimensional phenomic parameters as well as the mechanisms underlying the correlations; and third, phenomics-based studies are big data-driven studies, which can significantly enhance the possibility and efficiency for generating novel discoveries. However, phenomic studies on human diseases are still in early developmental stage, which are facing multiple major challenges and tasks: first, there is significant deficiency in analytical and modeling approaches for analyzing the multi-dimensional data of human phenomes; second, it is crucial to establish universal standards for acquirement and management of phenomic data of patients; third, new methods and devices for acquirement of phenomic data of patients under clinical settings should be developed; fourth, it is of significance to establish the regulatory and ethical guidelines for phenomic studies on diseases; and fifth, it is important to develop effective international cooperation. It is expected that phenomic studies on diseases would profoundly and comprehensively enhance our capacity in prevention, diagnosis and treatment of diseases.

Development and validation of explainable machine-learning models for carotid atherosclerosis early screening

BACKGROUND

Carotid atherosclerosis (CAS), an important factor in the development of stroke, is a major public health concern. The aim of this study was to establish and validate machine learning (ML) models for early screening of CAS using routine health check-up indicators in northeast China.

METHODS

A total of 69,601 health check-up records from the health examination center of the First Hospital of China Medical University (Shenyang, China) were collected between 2018 and 2019. For the 2019 records, 80% were assigned to the training set and 20% to the testing set. The 2018 records were used as the external validation dataset. Ten ML algorithms, including decision tree (DT), K-nearest neighbors (KNN), logistic regression (LR), naive Bayes (NB), random forest (RF), multiplayer perceptron (MLP), extreme gradient boosting machine (XGB), gradient boosting decision tree (GBDT), linear support vector machine (SVM-linear), and non-linear support vector machine (SVM-nonlinear), were used to construct CAS screening models. The area under the receiver operating characteristic curve (auROC) and precision-recall curve (auPR) were used as measures of model performance. The SHapley Additive exPlanations (SHAP) method was used to demonstrate the interpretability of the optimal model.

RESULTS

A total of 6315 records of patients undergoing carotid ultrasonography were collected; of these, 1632, 407, and 1141 patients were diagnosed with CAS in the training, internal validation, and external validation datasets, respectively. The GBDT model achieved the highest performance metrics with auROC of 0.860 (95% CI 0.839-0.880) in the internal validation dataset and 0.851 (95% CI 0.837-0.863) in the external validation dataset. Individuals with diabetes or those over 65 years of age showed low negative predictive value. In the interpretability analysis, age was the most important factor influencing the performance of the GBDT model, followed by sex and non-high-density lipoprotein cholesterol.

CONCLUSIONS

The ML models developed could provide good performance for CAS identification using routine health check-up indicators and could hopefully be applied in scenarios without ethnic and geographic heterogeneity for CAS prevention.

Predicting the toxicity of nanoparticles using artificial intelligence tools: a systematic review

Nanoparticles have been used extensively in different scientific fields. Due to the possible destructive effects of nanoparticles on the environment or the biological systems, their toxicity evaluation is a crucial phase for studying nanomaterial safety. In the meantime, experimental approaches for toxicity assessment of various nanoparticles are expensive and time-consuming. Thus, an alternative technique, such as artificial intelligence (AI), could be valuable for predicting nanoparticle toxicity. Therefore, in this review, the AI tools were investigated for the toxicity assessment of nanomaterials. To this end, a systematic search was performed on PubMed, Web of Science, and Scopus databases. Articles were included or excluded based on pre-defined inclusion and exclusion criteria, and duplicate studies were excluded. Finally, twenty-six studies were included. The majority of the studies were conducted on metal oxide and metallic nanoparticles. In addition, Random Forest (RF) and Support Vector Machine (SVM) had the most frequency in the included studies. Most of the models demonstrated acceptable performance. Overall, AI could provide a robust, fast, and low-cost tool for the evaluation of nanoparticle toxicity.

Natural Language Processing in Electronic Health Records in relation to healthcare decision-making: A systematic review

BACKGROUND

Natural Language Processing (NLP) is widely used to extract clinical insights from Electronic Health Records (EHRs). However, the lack of annotated data, automated tools, and other challenges hinder the full utilisation of NLP for EHRs. Various Machine Learning (ML), Deep Learning (DL) and NLP techniques are studied and compared to understand the limitations and opportunities in this space comprehensively.

METHODOLOGY

After screening 261 articles from 11 databases, we included 127 papers for full-text review covering seven categories of articles: (1) medical note classification, (2) clinical entity recognition, (3) text summarisation, (4) deep learning (DL) and transfer learning architecture, (5) information extraction, (6) Medical language translation and (7) other NLP applications. This study follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

RESULT AND DISCUSSION

EHR was the most commonly used data type among the selected articles, and the datasets were primarily unstructured. Various ML and DL methods were used, with prediction or classification being the most common application of ML or DL. The most common use cases were: the International Classification of Diseases, Ninth Revision (ICD-9) classification, clinical note analysis, and named entity recognition (NER) for clinical descriptions and research on psychiatric disorders.

CONCLUSION

We find that the adopted ML models were not adequately assessed. In addition, the data imbalance problem is quite important, yet we must find techniques to address this underlining problem. Future studies should address key limitations in studies, primarily identifying Lupus Nephritis, Suicide Attempts, perinatal self-harmed and ICD-9 classification.

Comparing risk prediction models aimed at predicting hospitalizations for adverse drug events in community dwelling older adults: a protocol paper

Background

The objective of this paper is to describe the creation, validation, and comparison of two risk prediction modeling approaches for community-dwelling older adults to identify individuals at highest risk for adverse drug event-related hospitalizations. One approach will use traditional statistical methods, the second will use a machine learning approach.

Methods

We will construct medication, clinical, health care utilization, and other variables known to be associated with adverse drug event-related hospitalizations. To create the cohort, we will include older adults (≥ 65 years of age) empaneled to a primary care physician within the Cedars-Sinai Health System primary care clinics with polypharmacy (≥ 5 medications) or at least 1 medication commonly implicated in ADEs (certain oral hypoglycemics, anti-coagulants, anti-platelets, and insulins). We will use a Fine-Gray Cox proportional hazards model for one risk modeling approach and DataRobot, a data science and analytics platform, to run and compare several widely used supervised machine learning algorithms, including Random Forest, Support Vector Machine, Extreme Gradient Boosting (XGBoost), Decision Tree, Naïve Bayes, and K-Nearest Neighbors. We will use a variety of metrics to compare model performance and to assess the risk of algorithmic bias.

Discussion

In conclusion, we hope to develop a pragmatic model that can be implemented in the primary care setting to risk stratify older adults to further optimize medication management.

Hospital selection framework for remote MCD patients based on fuzzy q-rung orthopair environment

This research proposes a novel mobile health-based hospital selection framework for remote patients with multi-chronic diseases based on wearable body medical sensors that use the Internet of Things. The proposed framework uses two powerful multi-criteria decision-making (MCDM) methods, namely fuzzy-weighted zero-inconsistency and fuzzy decision by opinion score method for criteria weighting and hospital ranking. The development of both methods is based on a Q-rung orthopair fuzzy environment to address the uncertainty issues associated with the case study in this research. The other MCDM issues of multiple criteria, various levels of significance and data variation are also addressed. The proposed framework comprises two main phases, namely identification and development. The first phase discusses the telemedicine architecture selected, patient dataset used and decision matrix integrated. The development phase discusses criteria weighting by q-ROFWZIC and hospital ranking by q-ROFDOSM and their sub-associated processes. Weighting results by q-ROFWZIC indicate that the time of arrival criterion is the most significant across all experimental scenarios with (0.1837, 0.183, 0.230, 0.276, 0.335) for (q = 1, 3, 5, 7, 10), respectively. Ranking results indicate that Hospital (H-4) is the best-ranked hospital in all experimental scenarios. Both methods were evaluated based on systematic ranking and sensitivity analysis, thereby confirming the validity of the proposed framework.

Prediction of type 2 diabetes using genome-wide polygenic risk score and metabolic profiles: A machine learning analysis of population-based 10-year prospective cohort study

BACKGROUND

Previous work on predicting type 2 diabetes by integrating clinical and genetic factors has mostly focused on the Western population. In this study, we use genome-wide polygenic risk score (gPRS) and serum metabolite data for type 2 diabetes risk prediction in the Asian population.

METHODS

Data of 1425 participants from the Korean Genome and Epidemiology Study (KoGES) Ansan-Ansung cohort were used in this study. For gPRS analysis, genotypic and clinical information from KoGES health examinee (n = 58,701) and KoGES cardiovascular disease association (n = 8105) sub-cohorts were included. Linkage disequilibrium analysis identified 239,062 genetic variants that were used to determine the gPRS, while the metabolites were selected using the Boruta algorithm. We used bootstrapped cross-validation to evaluate logistic regression and random forest (RF)-based machine learning models. Finally, associations of gPRS and selected metabolites with the values of homeostatic model assessment of beta-cell function (HOMA-B) and insulin resistance (HOMA-IR) were further estimated.

FINDINGS

During the follow-up period (8.3 ± 2.8 years), 331 participants (23.2%) were diagnosed with type 2 diabetes. The areas under the curves of the RF-based models were 0.844, 0.876, and 0.883 for the model using only demographic and clinical factors, model including the gPRS, and model with both gPRS and metabolites, respectively. Incorporation of additional parameters in the latter two models improved the classification by 11.7% and 4.2% respectively. While gPRS was significantly associated with HOMA-B value, most metabolites had a significant association with HOMA-IR value.

INTERPRETATION

Incorporating both gPRS and metabolite data led to enhanced type 2 diabetes risk prediction by capturing distinct etiologies of type 2 diabetes development. An RF-based model using clinical factors, gPRS, and metabolites predicted type 2 diabetes risk more accurately than the logistic regression-based model.

FUNDING

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MEST) (No. 2019M3E5D1A02070863 and 2022R1C1C1005458). This work was also supported by the 2020 Research Fund (1.200098.01) of UNIST (Ulsan National Institute of Science & Technology).

Identification of the risk factors of type 2 diabetes and its prediction using machine learning techniques

This study identified the risk factors for type 2 diabetes (T2D) and proposed a machine learning (ML) technique for predicting T2D. The risk factors for T2D were identified by multiple logistic regression (MLR) using p-value (p<0.05). Then, five ML-based techniques, including logistic regression, naïve Bayes, J48, multilayer perceptron, and random forest (RF) were employed to predict T2D. This study utilized two publicly available datasets, derived from the National Health and Nutrition Examination Survey, 2009-2010 and 2011-2012. About 4922 respondents with 387 T2D patients were included in 2009-2010 dataset, whereas 4936 respondents with 373 T2D patients were included in 2011-2012. This study identified six risk factors (age, education, marital status, SBP, smoking, and BMI) for 2009-2010 and nine risk factors (age, race, marital status, SBP, DBP, direct cholesterol, physical activity, smoking, and BMI) for 2011-2012. RF-based classifier obtained 95.9% accuracy, 95.7% sensitivity, 95.3% F-measure, and 0.946 area under the curve.

Real-world data: a brief review of the methods, applications, challenges and opportunities

Background

The increased adoption of the internet, social media, wearable devices, e-health services, and other technology-driven services in medicine and healthcare has led to the rapid generation of various types of digital data, providing a valuable data source beyond the confines of traditional clinical trials, epidemiological studies, and lab-based experiments.

Methods

We provide a brief overview on the type and sources of real-world data and the common models and approaches to utilize and analyze real-world data. We discuss the challenges and opportunities of using real-world data for evidence-based decision making This review does not aim to be comprehensive or cover all aspects of the intriguing topic on RWD (from both the research and practical perspectives) but serves as a primer and provides useful sources for readers who interested in this topic.

Results and Conclusions

Real-world hold great potential for generating real-world evidence for designing and conducting confirmatory trials and answering questions that may not be addressed otherwise. The voluminosity and complexity of real-world data also call for development of more appropriate, sophisticated, and innovative data processing and analysis techniques while maintaining scientific rigor in research findings, and attentions to data ethics to harness the power of real-world data.

Accuracy of Machine Learning Classification Models for the Prediction of Type 2 Diabetes Mellitus: A Systematic Survey and Meta-Analysis Approach

Soft-computing and statistical learning models have gained substantial momentum in predicting type 2 diabetes mellitus (T2DM) disease. This paper reviews recent soft-computing and statistical learning models in T2DM using a meta-analysis approach. We searched for papers using soft-computing and statistical learning models focused on T2DM published between 2010 and 2021 on three different search engines. Of 1215 studies identified, 34 with 136952 patients met our inclusion criteria. The pooled algorithm's performance was able to predict T2DM with an overall accuracy of 0.86 (95% confidence interval [CI] of [0.82, 0.89]). The classification of diabetes prediction was significantly greater in models with a screening and diagnosis (pooled proportion [95% CI] = 0.91 [0.74, 0.97]) when compared to models with nephropathy (pooled proportion = 0.48 [0.76, 0.89] to 0.88 [0.83, 0.91]). For the prediction of T2DM, the decision trees (DT) models had a pooled accuracy of 0.88 [95% CI: 0.82, 0.92], and the neural network (NN) models had a pooled accuracy of 0.85 [95% CI: 0.79, 0.89]. Meta-regression did not provide any statistically significant findings for the heterogeneous accuracy in studies with different diabetes predictions, sample sizes, and impact factors. Additionally, ML models showed high accuracy for the prediction of T2DM. The predictive accuracy of ML algorithms in T2DM is promising, mainly through DT and NN models. However, there is heterogeneity among ML models. We compared the results and models and concluded that this evidence might help clinicians interpret data and implement optimum models for their dataset for T2DM prediction.

Using Machine Learning Techniques and National Tuberculosis Surveillance Data to Predict Excess Growth in Genotyped Tuberculosis Clusters

The early identification of clusters of persons with tuberculosis (TB) that will grow to become outbreaks creates an opportunity for intervention in preventing future TB cases. We used surveillance data (2009-2018) from the United States, statistically derived definitions of unexpected growth, and machine-learning techniques to predict which clusters of genotype-matched TB cases are most likely to continue accumulating cases above expected growth within a 1-year follow-up period. We developed a model to predict which clusters are likely to grow on a training and testing data set that was generalizable to a validation data set. Our model showed that characteristics of clusters were more important than the social, demographic, and clinical characteristics of the patients in those clusters. For instance, the time between cases before unexpected growth was identified as the most important of our predictors. A faster accumulation of cases increased the probability of excess growth being predicted during the follow-up period. We have demonstrated that combining the characteristics of clusters and cases with machine learning can add to existing tools to help prioritize which clusters may benefit most from public health interventions. For example, consideration of an entire cluster, not only an individual patient, may assist in interrupting ongoing transmission.

Treatment Discontinuation Prediction in Patients With Diabetes Using a Ranking Model: Machine Learning Model Development

BACKGROUND

Treatment discontinuation (TD) is one of the major prognostic issues in diabetes care, and several models have been proposed to predict a missed appointment that may lead to TD in patients with diabetes by using binary classification models for the early detection of TD and for providing intervention support for patients. However, as binary classification models output the probability of a missed appointment occurring within a predetermined period, they are limited in their ability to estimate the magnitude of TD risk in patients with inconsistent intervals between appointments, making it difficult to prioritize patients for whom intervention support should be provided.

OBJECTIVE

This study aimed to develop a machine-learned prediction model that can output a TD risk score defined by the length of time until TD and prioritize patients for intervention according to their TD risk.

METHODS

This model included patients with diagnostic codes indicative of diabetes at the University of Tokyo Hospital between September 3, 2012, and May 17, 2014. The model was internally validated with patients from the same hospital from May 18, 2014, to January 29, 2016. The data used in this study included 7551 patients who visited the hospital after January 1, 2004, and had diagnostic codes indicative of diabetes. In particular, data that were recorded in the electronic medical records between September 3, 2012, and January 29, 2016, were used. The main outcome was the TD of a patient, which was defined as missing a scheduled clinical appointment and having no hospital visits within 3 times the average number of days between the visits of the patient and within 60 days. The TD risk score was calculated by using the parameters derived from the machine-learned ranking model. The prediction capacity was evaluated by using test data with the C-index for the performance of ranking patients, area under the receiver operating characteristic curve, and area under the precision-recall curve for discrimination, in addition to a calibration plot.

RESULTS

The means (95% confidence limits) of the C-index, area under the receiver operating characteristic curve, and area under the precision-recall curve for the TD risk score were 0.749 (0.655, 0.823), 0.758 (0.649, 0.857), and 0.713 (0.554, 0.841), respectively. The observed and predicted probabilities were correlated with the calibration plots.

CONCLUSIONS

A TD risk score was developed for patients with diabetes by combining a machine-learned method with electronic medical records. The score calculation can be integrated into medical records to identify patients at high risk of TD, which would be useful in supporting diabetes care and preventing TD.

Development and Evaluation of a Natural Language Processing Annotation Tool to Facilitate Phenotyping of Cognitive Status in Electronic Health Records: Diagnostic Study

BACKGROUND

Electronic health records (EHRs) with large sample sizes and rich information offer great potential for dementia research, but current methods of phenotyping cognitive status are not scalable.

OBJECTIVE

The aim of this study was to evaluate whether natural language processing (NLP)-powered semiautomated annotation can improve the speed and interrater reliability of chart reviews for phenotyping cognitive status.

METHODS

In this diagnostic study, we developed and evaluated a semiautomated NLP-powered annotation tool (NAT) to facilitate phenotyping of cognitive status. Clinical experts adjudicated the cognitive status of 627 patients at Mass General Brigham (MGB) health care, using NAT or traditional chart reviews. Patient charts contained EHR data from two data sets: (1) records from January 1, 2017, to December 31, 2018, for 100 Medicare beneficiaries from the MGB Accountable Care Organization and (2) records from 2 years prior to COVID-19 diagnosis to the date of COVID-19 diagnosis for 527 MGB patients. All EHR data from the relevant period were extracted; diagnosis codes, medications, and laboratory test values were processed and summarized; clinical notes were processed through an NLP pipeline; and a web tool was developed to present an integrated view of all data. Cognitive status was rated as cognitively normal, cognitively impaired, or undetermined. Assessment time and interrater agreement of NAT compared to manual chart reviews for cognitive status phenotyping was evaluated.

RESULTS

NAT adjudication provided higher interrater agreement (Cohen κ=0.89 vs κ=0.80) and significant speed up (time difference mean 1.4, SD 1.3 minutes; P<.001; ratio median 2.2, min-max 0.4-20) over manual chart reviews. There was moderate agreement with manual chart reviews (Cohen κ=0.67). In the cases that exhibited disagreement with manual chart reviews, NAT adjudication was able to produce assessments that had broader clinical consensus due to its integrated view of highlighted relevant information and semiautomated NLP features.

CONCLUSIONS

NAT adjudication improves the speed and interrater reliability for phenotyping cognitive status compared to manual chart reviews. This study underscores the potential of an NLP-based clinically adjudicated method to build large-scale dementia research cohorts from EHRs.

Methylation risk scores are associated with a collection of phenotypes within electronic health record systems

Inference of clinical phenotypes is a fundamental task in precision medicine, and has therefore been heavily investigated in recent years in the context of electronic health records (EHR) using a large arsenal of machine learning techniques, as well as in the context of genetics using polygenic risk scores (PRS). In this work, we considered the epigenetic analog of PRS, methylation risk scores (MRS), a linear combination of methylation states. We measured methylation across a large cohort (n = 831) of diverse samples in the UCLA Health biobank, for which both genetic and complete EHR data are available. We constructed MRS for 607 phenotypes spanning diagnoses, clinical lab tests, and medication prescriptions. When added to a baseline set of predictive features, MRS significantly improved the imputation of 139 outcomes, whereas the PRS improved only 22 (median improvement for methylation 10.74%, 141.52%, and 15.46% in medications, labs, and diagnosis codes, respectively, whereas genotypes only improved the labs at a median increase of 18.42%). We added significant MRS to state-of-the-art EHR imputation methods that leverage the entire set of medical records, and found that including MRS as a medical feature in the algorithm significantly improves EHR imputation in 37% of lab tests examined (median R² increase 47.6%). Finally, we replicated several MRS in multiple external studies of methylation (minimum p-value of 2.72 × 10⁻⁷) and replicated 22 of 30 tested MRS internally in two separate cohorts of different ethnicity. Our publicly available results and weights show promise for methylation risk scores as clinical and scientific tools.

Novel digital approaches to the assessment of problematic opioid use

The opioid epidemic continues to contribute to loss of life through overdose and significant social and economic burdens. Many individuals who develop problematic opioid use (POU) do so after being exposed to prescribed opioid analgesics. Therefore, it is important to accurately identify and classify risk factors for POU. In this review, we discuss the etiology of POU and highlight novel approaches to identifying its risk factors. These approaches include the application of polygenic risk scores (PRS) and diverse machine learning (ML) algorithms used in tandem with data from electronic health records (EHR), clinical notes, patient demographics, and digital footprints. The implementation and synergy of these types of data and approaches can greatly assist in reducing the incidence of POU and opioid-related mortality by increasing the knowledge base of patient-related risk factors, which can help to improve prescribing practices for opioid analgesics.

Application of machine learning methods for the prediction of true fasting status in patients performing blood tests

The fasting blood glucose (FBG) values extracted from electronic medical records (EMR) are assumed valid in existing research, which may cause diagnostic bias due to misclassification of fasting status. We proposed a machine learning (ML) algorithm to predict the fasting status of blood samples. This cross-sectional study was conducted using the EMR of a medical center from 2003 to 2018 and a total of 2,196,833 ontological FBGs from the outpatient service were enrolled. The theoretical true fasting status are identified by comparing the values of ontological FBG with average glucose levels derived from concomitant tested HbA1c based on multi-criteria. In addition to multiple logistic regression, we extracted 67 features to predict the fasting status by eXtreme Gradient Boosting (XGBoost). The discrimination and calibration of the prediction models were also assessed. Real-world performance was gauged by the prevalence of ineffective glucose measurement (IGM). Of the 784,340 ontologically labeled fasting samples, 77.1% were considered theoretical FBGs. The median (IQR) glucose and HbA1c level of ontological and theoretical fasting samples in patients without diabetes mellitus (DM) were 94.0 (87.0, 102.0) mg/dL and 5.6 (5.4, 5.9)%, and 92.0 (86.0, 99.0) mg/dL and 5.6 (5.4, 5.9)%, respectively. The XGBoost showed comparable calibration and AUROC of 0.887 than that of 0.868 in multiple logistic regression in the parsimonious approach and identified important predictors of glucose level, home-to-hospital distance, age, and concomitantly serum creatinine and lipid testing. The prevalence of IGM dropped from 27.8% based on ontological FBGs to 0.48% by using algorithm-verified FBGs. The proposed ML algorithm or multiple logistic regression model aids in verification of the fasting status.

Fallacy of Median Door‐to‐ECG Time: Hidden Opportunities for STEMI Screening Improvement

Background

ST‐segment elevation myocardial infarction (STEMI) guidelines recommend screening arriving emergency department (ED) patients for an early ECG in those with symptoms concerning for myocardial ischemia. Process measures target median door‐to‐ECG (D2E) time of 10 minutes.

Methods and Results

This 3‐year descriptive retrospective cohort study, including 676 ED‐diagnosed patients with STEMI from 10 geographically diverse facilities across the United States, examines an alternative approach to quantifying performance: proportion of patients meeting the goal of D2E≤10 minutes. We also identified characteristics associated with D2E>10 minutes and estimated the proportion of patients with screening ECG occurring during intake, triage, and main ED care periods. We found overall median D2E was 7 minutes (IQR:4–16; range: 0–1407 minutes; range of ED medians: 5–11 minutes). Proportion of patients with D2E>10 minutes was 37.9% (ED range: 21.5%–57.1%). Patients with D2E>10 minutes, compared to those with D2E≤10 minutes, were more likely female (32.8% versus 22.6%, P=0.005), Black (23.4% versus 12.4%, P=0.005), non‐English speaking (24.6% versus 19.5%, P=0.032), diabetic (40.2% versus 30.2%, P=0.010), and less frequently reported chest pain (63.3% versus 87.4%, P<0.001). ECGs were performed during ED intake in 62.1% of visits, ED triage in 25.3%, and main ED care in 12.6%.

Conclusions

Examining D2E>10 minutes can identify opportunities to improve care for more ED patients with STEMI. Our findings suggest sex, race, language, and diabetes are associated with STEMI diagnostic delays. Moving the acquisition of ECGs completed during triage to intake could achieve the D2E≤10 minutes goal for 87.4% of ED patients with STEMI. Sophisticated screening, accounting for differential risk and diversity in STEMI presentations, may further improve timely detection.

Application of unsupervised deep learning algorithms for identification of specific clusters of chronic cough patients from EMR data

BACKGROUND

Chronic cough affects approximately 10% of adults. The lack of ICD codes for chronic cough makes it challenging to apply supervised learning methods to predict the characteristics of chronic cough patients, thereby requiring the identification of chronic cough patients by other mechanisms. We developed a deep clustering algorithm with auto-encoder embedding (DCAE) to identify clusters of chronic cough patients based on data from a large cohort of 264,146 patients from the Electronic Medical Records (EMR) system. We constructed features using the diagnosis within the EMR, then built a clustering-oriented loss function directly on embedded features of the deep autoencoder to jointly perform feature refinement and cluster assignment. Lastly, we performed statistical analysis on the identified clusters to characterize the chronic cough patients compared to the non-chronic cough patients.

RESULTS

The experimental results show that the DCAE model generated three chronic cough clusters and one non-chronic cough patient cluster. We found various diagnoses, medications, and lab tests highly associated with chronic cough patients by comparing the chronic cough cluster with the non-chronic cough cluster. Comparison of chronic cough clusters demonstrated that certain combinations of medications and diagnoses characterize some chronic cough clusters.

CONCLUSIONS

To the best of our knowledge, this study is the first to test the potential of unsupervised deep learning methods for chronic cough investigation, which also shows a great advantage over existing algorithms for patient data clustering.

Machine learning for diabetes clinical decision support: a review

Type 2 diabetes has recently acquired the status of an epidemic silent killer, though it is non-communicable. There are two main reasons behind this perception of the disease. First, a gradual but exponential growth in the disease prevalence has been witnessed irrespective of age groups, geography or gender. Second, the disease dynamics are very complex in terms of multifactorial risks involved, initial asymptomatic period, different short-term and long-term complications posing serious health threat and related co-morbidities. Majority of its risk factors are lifestyle habits like physical inactivity, lack of exercise, high body mass index (BMI), poor diet, smoking except some inevitable ones like family history of diabetes, ethnic predisposition, ageing etc. Nowadays, machine learning (ML) is increasingly being applied for alleviation of diabetes health burden and many research works have been proposed in the literature to offer clinical decision support in different application areas as well. In this paper, we present a review of such efforts for the prevention and management of type 2 diabetes. Firstly, we present the medical gaps in diabetes knowledge base, guidelines and medical practice identified from relevant articles and highlight those that can be addressed by ML. Further, we review the ML research works in three different application areas namely—(1) risk assessment (statistical risk scores and ML-based risk models), (2) diagnosis (using non-invasive and invasive features), (3) prognosis (from normoglycemia/prior morbidity to incident diabetes and prognosis of incident diabetes to related complications). We discuss and summarize the shortcomings or gaps in the existing ML methodologies for diabetes to be addressed in future. This review provides the breadth of ML predictive modeling applications for diabetes while highlighting the medical and technological gaps as well as various aspects involved in ML-based diabetes clinical decision support.

Benchmarking missing-values approaches for predictive models on health databases

Background

As databases grow larger, it becomes harder to fully control their collection, and they frequently come with missing values. These large databases are well suited to train machine learning models, e.g., for forecasting or to extract biomarkers in biomedical settings. Such predictive approaches can use discriminative—rather than generative—modeling and thus open the door to new missing-values strategies. Yet existing empirical evaluations of strategies to handle missing values have focused on inferential statistics.

Results

Here we conduct a systematic benchmark of missing-values strategies in predictive models with a focus on large health databases: 4 electronic health record datasets, 1 population brain imaging database, 1 health survey, and 2 intensive care surveys. Using gradient-boosted trees, we compare native support for missing values with simple and state-of-the-art imputation prior to learning. We investigate prediction accuracy and computational time. For prediction after imputation, we find that adding an indicator to express which values have been imputed is important, suggesting that the data are missing not at random. Elaborate missing-values imputation can improve prediction compared to simple strategies but requires longer computational time on large data. Learning trees that model missing values—with missing incorporated attribute—leads to robust, fast, and well-performing predictive modeling.

Conclusions

Native support for missing values in supervised machine learning predicts better than state-of-the-art imputation with much less computational cost. When using imputation, it is important to add indicator columns expressing which values have been imputed.

Use of Machine Learning and Routine Laboratory Tests for Diabetes Mellitus Screening

Most patients with diabetes mellitus are asymptomatic, which leads to delayed and more complex treatment. At the same time, most individuals are routinely subjected to standard clinical laboratory examinations, which create large health datasets over a lifetime. Computer processing has been used to search for health anomalies and predict diseases using clinical examinations. This work studied machine learning models to support the screening of diabetes through routine laboratory tests using data from laboratory tests of 62,496 patients. The classification and regression models used were the K-nearest neighbor, support vector machines, Bayes naïve, random forest models, and artificial neural networks. Glycated hemoglobin, a test used for diabetes diagnosis, was used as the target. Regression models calculated glycated hemoglobin directly and were later classified. The performance of classification computer models has been studied under various subdataset partitions and combinations (e.g., healthy, prediabetic, and diabetes, as well as no healthy and no diabetes). The best single performance was achieved with the artificial neural network model when detecting prediabetes or diabetes. The artificial neural network classification model scored 78.1%, 78.7%, and 78.4% for sensitivity, precision, and F1 scores, respectively, when identifying no healthy group. Other models also had good results, depending on what is desired. Machine learning-based models can predict glycated hemoglobin values from routine laboratory tests and can be used as a screening tool to refer a patient for further testing.

Development of a visual attention based decision support system for autism spectrum disorder screening

Visual attention of young children with autism spectrum disorder (ASD) has been well documented in the literature for the past 20 years. In this study, we developed a Decision Support System (DSS) that uses machine learning (ML) techniques to identify young children with ASD from typically developing (TD) children. Study participants included 26 to 36 months old young children with ASD (n = 61) and TD children (n = 72). The results showed that the proposed DSS achieved up to 87.5% success rate in the early assessment of ASD in young children. Findings suggested that visual attention is a unique, promising biomarker for early assessment of ASD. Study results were discussed, and suggestions for future research were provided.

Methodological guidelines to estimate population-based health indicators using linked data and/or machine learning techniques

BACKGROUND

The capacity to use data linkage and artificial intelligence to estimate and predict health indicators varies across European countries. However, the estimation of health indicators from linked administrative data is challenging due to several reasons such as variability in data sources and data collection methods resulting in reduced interoperability at various levels and timeliness, availability of a large number of variables, lack of skills and capacity to link and analyze big data. The main objective of this study is to develop the methodological guidelines calculating population-based health indicators to guide European countries using linked data and/or machine learning (ML) techniques with new methods.

METHOD

We have performed the following step-wise approach systematically to develop the methodological guidelines: i. Scientific literature review, ii. Identification of inspiring examples from European countries, and iii. Developing the checklist of guidelines contents.

RESULTS

We have developed the methodological guidelines, which provide a systematic approach for studies using linked data and/or ML-techniques to produce population-based health indicators. These guidelines include a detailed checklist of the following items: rationale and objective of the study (i.e., research question), study design, linked data sources, study population/sample size, study outcomes, data preparation, data analysis (i.e., statistical techniques, sensitivity analysis and potential issues during data analysis) and study limitations.

CONCLUSIONS

This is the first study to develop the methodological guidelines for studies focused on population health using linked data and/or machine learning techniques. These guidelines would support researchers to adopt and develop a systematic approach for high-quality research methods. There is a need for high-quality research methodologies using more linked data and ML-techniques to develop a structured cross-disciplinary approach for improving the population health information and thereby the population health.

Medication based machine learning to identify subpopulations of pediatric hemodialysis patients in an electronic health record database

Electronic health records (EHRs) have given rise to large and complex databases of medical information that have the potential to become powerful tools for clinical research. However, differences in coding systems and the detail and accuracy of the information within EHRs can vary across institutions. This makes it challenging to identify subpopulations of patients and limits the widespread use of multi-institutional databases. In this study, we leveraged machine learning to identify patterns in medication usage among hospitalized pediatric patients receiving renal replacement therapy and created a predictive model that successfully differentiated between intermittent (iHD) and continuous renal replacement therapy (CRRT) hemodialysis patients. We trained six machine learning algorithms (logistical regression, Naïve Bayes, k-nearest neighbor, support vector machine, random forest, and gradient boosted trees) using patient records from a multi-center database (n = 533) and prescribed medication ingredients (n = 228) as features to discriminate between the two hemodialysis types. Predictive skill was assessed using a 5-fold cross-validation, and the algorithms showed a range of performance from 0.7 balanced accuracy (logistical regression) to 0.86 (random forest). The two best performing models were further tested using an independent single-center dataset and achieved 84–87% balanced accuracy. This model overcomes issues inherent within large databases and will allow us to utilize and combine historical records, significantly increasing population size and diversity within both iHD and CRRT populations for future clinical studies. Our work demonstrates the utility of using medications alone to accurately differentiate subpopulations of patients in large datasets, allowing codes to be transferred between different coding systems. This framework has the potential to be used to distinguish other subpopulations of patients where discriminatory ICD codes are not available, permitting more detailed insights and new lines of research.

Characterization of Symptoms and Symptom Clusters for Type 2 Diabetes Using a Large Nationwide Electronic Health Record Database

OBJECTIVE

A variety of symptoms may be associated with type 2 diabetes and its complications. Symptoms in chronic diseases may be described in terms of prevalence, severity, and trajectory and often co-occur in groups, known as symptom clusters, which may be representative of a common etiology. The purpose of this study was to characterize type 2 diabetes-related symptoms using a large nationwide electronic health record (EHR) database.

METHODS

We acquired the Cerner Health Facts, a nationwide EHR database. The type 2 diabetes cohort (n = 1,136,301 patients) was identified using a rule-based phenotype method. A multistep procedure was then used to identify type 2 diabetes-related symptoms based on International Classification of Diseases, 9th and 10th revisions, diagnosis codes. Type 2 diabetes-related symptoms and co-occurring symptom clusters, including their temporal patterns, were characterized based the longitudinal EHR data.

RESULTS

Patients had a mean age of 61.4 years, 51.2% were female, and 70.0% were White. Among 1,136,301 patients, there were 8,008,276 occurrences of 59 symptoms. The most frequently reported symptoms included pain, heartburn, shortness of breath, fatigue, and swelling, which occurred in 21-60% of the patients. We also observed over-represented type 2 diabetes symptoms, including difficulty speaking, feeling confused, trouble remembering, weakness, and drowsiness/sleepiness. Some of these are rare and difficult to detect by traditional patient-reported outcomes studies.

CONCLUSION

To the best of our knowledge, this is the first study to use a nationwide EHR database to characterize type 2 diabetes-related symptoms and their temporal patterns. Fifty-nine symptoms, including both over-represented and rare diabetes-related symptoms, were identified.

Different Data Mining Approaches Based Medical Text Data

The amount of medical text data is increasing dramatically. Medical text data record the progress of medicine and imply a large amount of medical knowledge. As a natural language, they are characterized by semistructured, high-dimensional, high data volume semantics and cannot participate in arithmetic operations. Therefore, how to extract useful knowledge or information from the total available data is very important task. Using various techniques of data mining can extract valuable knowledge or information from data. In the current study, we reviewed different approaches to apply for medical text data mining. The advantages and shortcomings for each technique compared to different processes of medical text data were analyzed. We also explored the applications of algorithms for providing insights to the users and enabling them to use the resources for the specific challenges in medical text data. Further, the main challenges in medical text data mining were discussed. Findings of this paper are benefit for helping the researchers to choose the reasonable techniques for mining medical text data and presenting the main challenges to them in medical text data mining.

A comprehensive review of machine learning techniques on diabetes detection

Diabetes mellitus has been an increasing concern owing to its high morbidity, and the average age of individual affected by of individual affected by this disease has now decreased to mid-twenties. Given the high prevalence, it is necessary to address with this problem effectively. Many researchers and doctors have now developed detection techniques based on artificial intelligence to better approach problems that are missed due to human errors. Data mining techniques with algorithms such as - density-based spatial clustering of applications with noise and ordering points to identify the cluster structure, the use of machine vision systems to learn data on facial images, gain better features for model training, and diagnosis via presentation of iridocyclitis for detection of the disease through iris patterns have been deployed by various practitioners. Machine learning classifiers such as support vector machines, logistic regression, and decision trees, have been comparative discussed various authors. Deep learning models such as artificial neural networks and recurrent neural networks have been considered, with primary focus on long short-term memory and convolutional neural network architectures in comparison with other machine learning models. Various parameters such as the root-mean-square error, mean absolute errors, area under curves, and graphs with varying criteria are commonly used. In this study, challenges pertaining to data inadequacy and model deployment are discussed. The future scope of such methods has also been discussed, and new methods are expected to enhance the performance of existing models, allowing them to attain greater insight into the conditions on which the prevalence of the disease depends.

Estimation of Heart Rate Using Regression Models and Artificial Neural Network in Middle-Aged Adults

Purpose: Heart rate is the most commonly used indicator in clinical medicine to assess the functionality of the cardiovascular system. Most studies have focused on age-based equations to estimate the maximal heart rate, neglecting multiple factors that affect the accuracy of the prediction. Methods: We studied 121 middle-aged adults at an average age of 57.2years with an average body mass index (BMI) of 25.9. The participants performed on a power bike with a starting wattage of 0W that was increased by 25W every 3min until the experiment terminated. Ambulatory blood pressure and electrocardiography were monitored through gas metabolic analyzers for safety concerns. Six descriptive characteristics of participants were observed, which were further analyzed using a multivariate regression model and an artificial neural network (ANN). Results: The input variables for the multivariate regression model and ANN were selected by correlation for the reduction of dimension. The accuracy of estimation by multivariate regression model and ANN was 9.74 and 9.42%, respectively, which outperformed the traditional age-based model (with an accuracy of 10.31%). Conclusion: This study provides comprehensive approaches to estimate the maximal heart rate using multiple indicators, revealing that both the multivariate regression model and ANN incorporated with age, resting heart rate (RHR), and second-order heart rate (SOHR) are more accurate than univariate models.

Applying interpretable deep learning models to identify chronic cough patients using EHR data

BACKGROUND AND OBJECTIVE

Chronic cough (CC) affects approximately 10% of adults. Many disease states are associated with chronic cough, such as asthma, upper airway cough syndrome, bronchitis, and gastroesophageal reflux disease. The lack of an ICD code specific for chronic cough makes it challenging to identify such patients from electronic health records (EHRs). For clinical and research purposes, computational methods using EHR data are urgently needed to identify chronic cough cases. This research aims to investigate the data representations and deep learning algorithms for chronic cough prediction.

METHODS

Utilizing real-world EHR data from a large academic healthcare system from October 2005 to September 2015, we investigated Natural Language Representation of the EHR data and systematically evaluated deep learning and traditional machine learning models to predict chronic cough patients. We built these machine learning models using structured data (medication and diagnosis) and unstructured data (clinical notes).

RESULTS

The sensitivity and specificity of a transformer-based deep learning algorithm, specifically BERT with attention model, was 0.856 and 0.866, respectively, using structured data (medication and diagnosis). Sensitivity and specificity improved to 0.952 and 0.930 when we combined structured data with symptoms extracted from clinical notes. We further found that the attention mechanism of deep learning models can be used to extract important features that drive the prediction decisions. Compared with our previously published rule-based algorithm, the deep learning algorithm can identify more chronic cough patients with structured data.

CONCLUSIONS

By applying deep learning models, chronic cough patients can be reliably identified for prospective or retrospective research through medication and diagnosis data, widely available in EHR and electronic claims data, thus improving the generalizability of the patient identification algorithm. Deep learning models can identify chronic cough patients with even higher sensitivity and specificity when structured and unstructured EHR data are utilized. We anticipate language-based data representation and deep learning models developed in this research could also be productively used for other disease prediction and case identification.

High-throughput phenotyping with temporal sequences

OBJECTIVE

High-throughput electronic phenotyping algorithms can accelerate translational research using data from electronic health record (EHR) systems. The temporal information buried in EHRs is often underutilized in developing computational phenotypic definitions. This study aims to develop a high-throughput phenotyping method, leveraging temporal sequential patterns from EHRs.

MATERIALS AND METHODS

We develop a representation mining algorithm to extract 5 classes of representations from EHR diagnosis and medication records: the aggregated vector of the records (aggregated vector representation), the standard sequential patterns (sequential pattern mining), the transitive sequential patterns (transitive sequential pattern mining), and 2 hybrid classes. Using EHR data on 10 phenotypes from the Mass General Brigham Biobank, we train and validate phenotyping algorithms.

RESULTS

Phenotyping with temporal sequences resulted in a superior classification performance across all 10 phenotypes compared with the standard representations in electronic phenotyping. The high-throughput algorithm's classification performance was superior or similar to the performance of previously published electronic phenotyping algorithms. We characterize and evaluate the top transitive sequences of diagnosis records paired with the records of risk factors, symptoms, complications, medications, or vaccinations.

DISCUSSION

The proposed high-throughput phenotyping approach enables seamless discovery of sequential record combinations that may be difficult to assume from raw EHR data. Transitive sequences offer more accurate characterization of the phenotype, compared with its individual components, and reflect the actual lived experiences of the patients with that particular disease.

CONCLUSION

Sequential data representations provide a precise mechanism for incorporating raw EHR records into downstream machine learning. Our approach starts with user interpretability and works backward to the technology.