Random forest approach for determining risk prediction and predictive factors of type 2 diabetes: large-scale health check-up data in Japan

Strong Effect of Demographic Changes on Tuberculosis Susceptibility in South Africa

South Africa is among the world's top eight TB burden countries, and despite a focus on HIV-TB co-infection, most of the population living with TB are not HIV co-infected. The disease is endemic across the country with 80-90% exposure by adulthood. We investigated epidemiological risk factors for tuberculosis (TB) in the Northern Cape Province, South Africa: an understudied TB endemic region with extreme TB incidence (645/100,000) and the lowest provincial population density. We leveraged the population's high TB incidence and community transmission to design a case-control study with population-based controls, reflecting similar mechanisms of exposure between the groups. We recruited 1,126 participants with suspected TB from 12 community health clinics, and generated a cohort of 878 individuals (cases =374, controls =504) after implementing our enrollment criteria. All participants were GeneXpert Ultra tested for active TB by a local clinic. We assessed important risk factors for active TB using logistic regression and random forest modeling. Additionally, a subset of individuals were genotyped to determine genome-wide ancestry components. Male gender had the strongest effect on TB risk (OR: 2.87 [95% CI: 2.1-3.8]); smoking and alcohol consumption did not significantly increase TB risk. We identified two interactions: age by socioeconomic status (SES) and birthplace by residence locality on TB risk (OR = 3.05, p = 0.016) - where rural birthplace but town residence was the highest risk category. Finally, participants had a majority Khoe-San ancestry, typically greater than 50%. Epidemiological risk factors for this cohort differ from other global populations. The significant interaction effects reflect rapid changes in SES and mobility over recent generations and strongly impact TB risk in the Northern Cape of South Africa. Our models show that such risk factors combined explain 16% of the variance (r2) in case/control status.

Machine learning risk estimation and prediction of death in continuing care facilities using administrative data

In this study, we aimed to identify the factors that were associated with mortality among continuing care residents in Alberta, during the coronavirus disease 2019 (COVID-19) pandemic. We achieved this by leveraging and linking various administrative datasets together. Then, we examined pre-processing methods in terms of prediction performance. Finally, we developed several machine learning models and compared the results of these models in terms of performance. We conducted a retrospective cohort study of all continuing care residents in Alberta, Canada, from March 1, 2020, to March 31, 2021. We used a univariable and a multivariable logistic regression (LR) model to identify predictive factors of 60-day all-cause mortality by estimating odds ratios (ORs) with a 95% confidence interval. To determine the best sensitivity-specificity cut-off point, the Youden index was employed. We developed several machine learning models to determine the best model regarding performance. In this cohort study, increased age, male sex, symptoms, previous admissions, and some specific comorbidities were associated with increased mortality. Machine learning and pre-processing approaches offer a potentially valuable method for improving risk prediction for mortality, but more work is needed to show improvement beyond standard risk factors.

Construction and validation of a predictive model for postoperative urinary retention after lumbar interbody fusion surgery

BACKGROUND

Postoperative urine retention (POUR) after lumbar interbody fusion surgery may lead to recatheterization and prolonged hospitalization. In this study, a predictive model was constructed and validated. The objective was to provide a nomogram for estimating the risk of POUR and then reducing the incidence.

METHODS

A total of 423 cases of lumbar fusion surgery were included; 65 of these cases developed POUR, an incidence of 15.4%. The dataset is divided into a training set and a validation set according to time. 18 candidate variables were selected. The candidate variables were screened through LASSO regression. The stepwise regression and random forest analysis were then conducted to construct the predictive model and draw a nomogram. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve and the calibration curve were used to evaluate the predictive effect of the model.

RESULTS

The best lambda value in LASSO was 0.025082; according to this, five significant variables were screened, including age, smoking history, surgical method, operative time, and visual analog scale (VAS) score of postoperative low back pain. A predictive model containing four variables was constructed by stepwise regression. The variables included age (β = 0.047, OR = 1.048), smoking history (β = 1.950, OR = 7.031), operative time (β = 0.022, OR = 1.022), and postoperative VAS score of low back pain (β = 2.554, OR = 12.858). A nomogram was drawn based on the results. The AUC of the ROC curve of the training set was 0.891, the validation set was 0.854 in the stepwise regression model. The calibration curves of the training set and validation set are in good agreement with the actual curves, showing that the stepwise regression model has good prediction ability. The AUC of the training set was 0.996, and that of the verification set was 0.856 in the random forest model.

CONCLUSION

This study developed and internally validated a new nomogram and a random forest model for predicting the risk of POUR after lumbar interbody fusion surgery. Both of the nomogram and the random forest model have high accuracy in this study.

Developing a prediction model of children asthma risk using population-based family history health records

BACKGROUND

Identifying children at high risk of developing asthma can facilitate prevention and early management strategies. We developed a prediction model of children's asthma risk using objectively collected population-based children and parental histories of comorbidities.

METHODS

We conducted a retrospective population-based cohort study using administrative data from Manitoba, Canada, and included children born from 1974 to 2000 with linkages to ≥1 parent. We identified asthma and prior comorbid condition diagnoses from hospital and outpatient records. We used two machine-learning models: least absolute shrinkage and selection operator (LASSO) logistic regression (LR) and random forest (RF) to identify important predictors. The predictors in the base model included children's demographics, allergic conditions, respiratory infections, and parental asthma. Subsequent models included additional multiple comorbidities for children and parents.

RESULTS

The cohort included 195,666 children: 51.3% were males and 17.7% had asthma diagnosis. The base LR model achieved a low predictive performance with sensitivity of 0.47, 95% confidence interval (0.45-0.48), and specificity of 0.67 (0.66-0.67) using a predicted probability threshold of 0.20. Sensitivity significantly improved when children's comorbidities were included using LASSO LR: 0.71 (0.69-0.72). Predictive performance further improved by including parental comorbidities (sensitivity = 0.72 [0.70-0.73], specificity = 0.69 [0.69-0.70]). We observed similar results for the RF models. Children's menstrual disorders and mood and anxiety disorders, parental lipid metabolism disorders and asthma were among the most important variables that predicted asthma risk.

CONCLUSION

Including children and parental comorbidities to children's asthma prediction models improves their accuracy.

CKD Progression Prediction in a Diverse US Population: A Machine-Learning Model

Rationale & Objective

Chronic kidney disease (CKD) is a major cause of morbidity and mortality. To date, there are no widely used machine-learning models that can predict progressive CKD across the entire disease spectrum, including the earliest stages. The objective of this study was to use readily available demographic and laboratory data from Sonic Healthcare USA laboratories to train and test the performance of machine learning-based predictive risk models for CKD progression.

Study Design

Retrospective observational study.

Setting & Participants

The study population was composed of deidentified laboratory information services data procured from a large US outpatient laboratory network. The retrospective data set included 110,264 adult patients over a 5-year period with initial estimated glomerular filtration rate (eGFR) values between 15-89 mL/min/1.73 m2.

Predictors

Patient demographic and laboratory characteristics.

Outcomes

Accelerated (ie, >30%) eGFR decline associated with CKD progression within 5 years.

Analytical Approach

Machine-learning models were developed using random forest survival methods, with laboratory-based risk factors analyzed as potential predictors of significant eGFR decline.

Results

The 7-variable risk classifier model accurately predicted an eGFR decline of >30% within 5 years and achieved an area under the curve receiver-operator characteristic of 0.85. The most important predictor of progressive decline in kidney function was the eGFR slope. Other key contributors to the model included initial eGFR, urine albumin-creatinine ratio, serum albumin (initial and slope), age, and sex.

Limitations

The cohort study did not evaluate the role of clinical variables (eg, blood pressure) on the performance of the model.

Conclusions

Our progressive CKD classifier accurately predicts significant eGFR decline in patients with early, mid, and advanced disease using readily obtainable laboratory data. Although prospective studies are warranted, our results support the clinical utility of the model to improve timely recognition and optimal management for patients at risk for CKD progression.

Plain-Language Summary

Defined by a significant decrease in estimated glomerular filtration rate (eGFR), chronic kidney disease (CKD) progression is strongly associated with kidney failure. However, to date, there are no broadly used resources that can predict this clinically significant event. Using machine-learning techniques on a diverse US population, this cohort study aimed to address this deficiency and found that a 5-year risk prediction model for CKD progression was accurate. The most important predictor of progressive decline in kidney function was the eGFR slope, followed by the urine albumin-creatinine ratio and serum albumin slope. Although further study is warranted, the results showed that a machine-learning model using readily obtainable laboratory information accurately predicts CKD progression, which may inform clinical diagnosis and management for this at-risk population.

Data collaboration analysis in predicting diabetes from a small amount of health checkup data

Recent studies showed that machine learning models such as gradient-boosting decision tree (GBDT) can predict diabetes with high accuracy from big data. In this study, we asked whether highly accurate prediction of diabetes is possible even from small data by expanding the amount of data through data collaboration (DC) analysis, a modern framework for integrating and analyzing data accumulated at multiple institutions while ensuring confidentiality. To this end, we focused on data from two institutions: health checkup data of 1502 citizens accumulated in Tsukuba City and health history data of 1399 patients collected at the University of Tsukuba Hospital. When using only the health checkup data, the ROC-AUC and Recall for logistic regression (LR) were 0.858 ± 0.014 and 0.970 ± 0.019, respectively, while those for GBDT were 0.856 ± 0.014 and 0.983 ± 0.016, respectively. When using also the health history data through DC analysis, these values for LR improved to 0.875 ± 0.013 and 0.993 ± 0.009, respectively, while those for GBDT deteriorated because of the low compatibility with a method used for confidential data sharing (although DC analysis brought improvements). Even in a situation where health checkup data of only 324 citizens are available, the ROC-AUC and Recall for LR were 0.767 ± 0.025 and 0.867 ± 0.04, respectively, thanks to DC analysis, indicating an 11% and 12% improvement. Thus, we concluded that the answer to the above question was "Yes" for LR but "No" for GBDT for the data set tested in this study.

Unleashing the Power of Very Small Data to Predict Acute Exacerbations of Chronic Obstructive Pulmonary Disease

Introduction

In this article, we explore to what extent it is possible to leverage on very small data to build machine learning (ML) models that predict acute exacerbations of chronic obstructive pulmonary disease (AECOPD).

Methods

We build ML models using the small data collected during the eHealth Diary telemonitoring study between 2013 and 2017 in Sweden. This data refers to a group of multimorbid patients, namely 18 patients with chronic obstructive pulmonary disease (COPD) as the major reason behind previous hospitalisations. The telemonitoring was supervised by a specialised hospital-based home care (HBHC) unit, which also was responsible for the medical actions needed.

Results

We implement two different ML approaches, one based on time-dependent covariates and the other one based on time-independent covariates. We compare the first approach with standard COX Proportional Hazards (CPH). For the second one, we use different proportions of synthetic data to build models and then evaluate the best model against authentic data.

Discussion

To the best of our knowledge, the present ML study shows for the first time that the most important variable for an increased risk of future AECOPDs is "maintenance medication changes by HBHC". This finding is clinically relevant since a sub-optimal maintenance treatment, requiring medication changes, puts the patient in risk for future AECOPDs.

Conclusion

The experiments return useful insights about the use of small data for ML.

Minimally invasive electrochemical continuous glucose monitoring sensors: Recent progress and perspective

Diabetes and its complications are seriously threatening the health and well-being of hundreds of millions of people. Glucose levels are essential indicators of the health conditions of diabetics. Over the past decade, concerted efforts in various fields have led to significant advances in glucose monitoring technology. In particular, the rapid development of continuous glucose monitoring (CGM) based on electrochemical sensing principles has great potential to overcome the limitations of self-monitoring blood glucose (SMBG) in continuously tracking glucose trends, evaluating diabetes treatment options, and improving the quality of life of diabetics. However, the applications of minimally invasive electrochemical CGM sensors are still limited owing to the following aspects: i) invasiveness, ii) short lifespan, iii) biocompatibility, and iv) calibration and prediction. In recent years, the performance of minimally invasive electrochemical CGM systems (CGMSs) has been significantly improved owing to breakthrough developments in new materials and key technologies. In this review, we summarize the history of commercial CGMSs, the development of sensing principles, and the research progress of minimally invasive electrochemical CGM sensors in reducing the invasiveness of implanted probes, maintaining enzyme activity, and improving the biocompatibility of the sensor interface. In addition, this review also introduces calibration algorithms and prediction algorithms applied to CGMSs and describes the application of machine learning algorithms for glucose prediction.

Artificial Intelligence (AI) and Internet of Medical Things (IoMT) Assisted Biomedical Systems for Intelligent Healthcare

Artificial intelligence (AI) is a modern approach based on computer science that develops programs and algorithms to make devices intelligent and efficient for performing tasks that usually require skilled human intelligence. AI involves various subsets, including machine learning (ML), deep learning (DL), conventional neural networks, fuzzy logic, and speech recognition, with unique capabilities and functionalities that can improve the performances of modern medical sciences. Such intelligent systems simplify human intervention in clinical diagnosis, medical imaging, and decision-making ability. In the same era, the Internet of Medical Things (IoMT) emerges as a next-generation bio-analytical tool that combines network-linked biomedical devices with a software application for advancing human health. In this review, we discuss the importance of AI in improving the capabilities of IoMT and point-of-care (POC) devices used in advanced healthcare sectors such as cardiac measurement, cancer diagnosis, and diabetes management. The role of AI in supporting advanced robotic surgeries developed for advanced biomedical applications is also discussed in this article. The position and importance of AI in improving the functionality, detection accuracy, decision-making ability of IoMT devices, and evaluation of associated risks assessment is discussed carefully and critically in this review. This review also encompasses the technological and engineering challenges and prospects for AI-based cloud-integrated personalized IoMT devices for designing efficient POC biomedical systems suitable for next-generation intelligent healthcare.

A Review of Feature Selection Methods for Machine Learning-Based Disease Risk Prediction

Machine learning has shown utility in detecting patterns within large, unstructured, and complex datasets. One of the promising applications of machine learning is in precision medicine, where disease risk is predicted using patient genetic data. However, creating an accurate prediction model based on genotype data remains challenging due to the so-called “curse of dimensionality” (i.e., extensively larger number of features compared to the number of samples). Therefore, the generalizability of machine learning models benefits from feature selection, which aims to extract only the most “informative” features and remove noisy “non-informative,” irrelevant and redundant features. In this article, we provide a general overview of the different feature selection methods, their advantages, disadvantages, and use cases, focusing on the detection of relevant features (i.e., SNPs) for disease risk prediction.

What is next for screening for undiagnosed atrial fibrillation? Artificial intelligence may hold the key

Atrial fibrillation (AF) is increasingly common, though often undiagnosed, leaving many people untreated and at elevated risk of ischaemic stroke. Current European guidelines do not recommend systematic screening for AF, even though a number of studies have shown that periods of serial or continuous rhythm monitoring in older people in the general population increase detection of AF and the prescription of oral anticoagulation. This article discusses the conflicting results of two contemporary landmark trials, STROKESTOP and the LOOP, which provided the first evidence on whether screening for AF confers a benefit for people in terms of clinical outcomes. The benefit and efficiency of systematic screening for AF in the general population could be optimized by targeting screening to only those at higher risk of developing AF. For this purpose, evidence is emerging that prediction models developed using artificial intelligence in routinely collected electronic health records can provide strong discriminative performance for AF and increase detection rates when combined with rhythm monitoring in a clinical study. We consider future directions for investigation in this field and how this could be best aligned to the current evidence base to target screening in people at elevated risk of stroke.

Predicting the Risk of Incident Type 2 Diabetes Mellitus in Chinese Elderly Using Machine Learning Techniques

Early identification of individuals at high risk of diabetes is crucial for implementing early intervention strategies. However, algorithms specific to elderly Chinese adults are lacking. The aim of this study is to build effective prediction models based on machine learning (ML) for the risk of type 2 diabetes mellitus (T2DM) in Chinese elderly. A retrospective cohort study was conducted using the health screening data of adults older than 65 years in Wuhan, China from 2018 to 2020. With a strict data filtration, 127,031 records from the eligible participants were utilized. Overall, 8298 participants were diagnosed with incident T2DM during the 2-year follow-up (2019–2020). The dataset was randomly split into training set (n = 101,625) and test set (n = 25,406). We developed prediction models based on four ML algorithms: logistic regression (LR), decision tree (DT), random forest (RF), and extreme gradient boosting (XGBoost). Using LASSO regression, 21 prediction features were selected. The Random under-sampling (RUS) was applied to address the class imbalance, and the Shapley Additive Explanations (SHAP) was used to calculate and visualize feature importance. Model performance was evaluated by the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and accuracy. The XGBoost model achieved the best performance (AUC = 0.7805, sensitivity = 0.6452, specificity = 0.7577, accuracy = 0.7503). Fasting plasma glucose (FPG), education, exercise, gender, and waist circumference (WC) were the top five important predictors. This study showed that XGBoost model can be applied to screen individuals at high risk of T2DM in the early phrase, which has the strong potential for intelligent prevention and control of diabetes. The key features could also be useful for developing targeted diabetes prevention interventions.

Commercial and Scientific Solutions for Blood Glucose Monitoring-A Review

Diabetes is a chronic and, according to the state of the art, an incurable disease. Therefore, to treat diabetes, regular blood glucose monitoring is crucial since it is mandatory to mitigate the risk and incidence of hyperglycemia and hypoglycemia. Nowadays, it is common to use blood glucose meters or continuous glucose monitoring via stinging the skin, which is classified as invasive monitoring. In recent decades, non-invasive monitoring has been regarded as a dominant research field. In this paper, electrochemical and electromagnetic non-invasive blood glucose monitoring approaches will be discussed. Thereby, scientific sensor systems are compared to commercial devices by validating the sensor principle and investigating their performance utilizing the Clarke error grid. Additionally, the opportunities to enhance the overall accuracy and stability of non-invasive glucose sensing and even predict blood glucose development to avoid hyperglycemia and hypoglycemia using post-processing and sensor fusion are presented. Overall, the scientific approaches show a comparable accuracy in the Clarke error grid to that of the commercial ones. However, they are in different stages of development and, therefore, need improvement regarding parameter optimization, temperature dependency, or testing with blood under real conditions. Moreover, the size of scientific sensing solutions must be further reduced for a wearable monitoring system.