Machine Learning Models Research Articles

Machine learning models predict the progression of long-term renal insufficiency in patients with renal cancer after radical nephrectomy

BackgroundChronic Kidney Disease (CKD) is a common severe complication after radical nephrectomy in patients with renal cancer. The timely and accurate prediction of the long-term progression of renal function post-surgery is crucial for early intervention and ultimately improving patient survival rates.ObjectiveThis study aimed to establish a machine learning model to predict the likelihood of long-term renal dysfunction progression after surgery by analyzing patients’ general information in depth.MethodsWe retrospectively collected data of eligible patients from the Affiliated Hospital of Qingdao University. The primary outcome was upgrading of the Chronic Kidney Disease stage between pre- and 3-year post-surgery. We constructed seven different machine-learning models based on Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (Lightgbm), Gaussian Naive Bayes (GaussianNB), and K-Nearest Neighbors (KNN). The performance of all predictive models was evaluated using the area under the receiver operating characteristic curve (AUC), precision-recall curves, confusion matrices, and calibration curves.ResultsAmong 360 patients with renal cancer who underwent radical nephrectomy included in this study, 185 (51.3%) experienced an upgrade in Chronic Kidney Disease stage 3-year post-surgery. Eleven predictive variables were selected for further construction of the machine learning models. The logistic regression model provided the most accurate prediction, with the highest AUC (0.8154) and an accuracy of 0.787.

Leveraging machine learning models for anemia severity detection among pregnant women following ANC: Ethiopian context

BackgroundAnemia during pregnancy is a significant public health concern, particularly in resource-limited settings. Machine learning (ML) offers promising avenues for improved anemia detection and management. This study investigates the potential of ML models in predicting anemia severity among pregnant women attending Antenatal Care (ANC) visits in Ethiopia.MethodsData from the Ethiopian Demographic Health Survey, specialized hospitals, and public hospitals were utilized. The dataset included individuals diagnosed with severe (65.12%), moderate (15.63%), mild (16.65%) anemia, and non-anemic (2.61%) cases. Feature selection employed filter methods based on mutual information, and F-score was used to assess anemia severity prediction across four classes. Six ML models (MLP-NN, XGBoost, GNB, Decision Tree, Random Forest, and KNN) were evaluated using accuracy, precision, recall, and F1-score.ResultsThe Random Forest classifier achieved the best overall performance across all categories, with an accuracy of 97%, precision of 93%, recall of 93%, and F1-score of 93%. This indicates high true positive rates and low false positive rates. While other models like XGBoost, MLP-NN, and Decision Tree showed good performance, they weren’t quite as strong as Random Forest. Classifiers like KNN and GNB had lower overall accuracy and a tendency to misclassify some cases.ConclusionsThis study demonstrates the promising potential of Random Forest in predicting anemia severity among pregnant women in Ethiopia. The findings contribute to a more holistic understanding of anemia risk factors and pave the way for improved early detection and targeted interventions.

Application of Machine Learning Models to Multi-Parameter Maximum Magnitude Prediction

Magnitude prediction is a key focus in earthquake science research, and using machine learning models to analyze seismic data, identify pre-seismic anomalies, and improve prediction accuracy is of great scientific and practical significance. Taking the southern part of China’s North–South Seismic Belt (20° N~30° N, 96° E~106° E), where strong earthquakes frequently occur, as an example, we used the sliding time window method to calculate 11 seismicity indicators from the earthquake catalog data as the characteristic parameters of the training model, and compared six machine learning models, including the random forest (RF) and long short-term memory (LSTM) models, to select the best-performing LSTM model for predicting the maximum magnitude of an earthquake in the study area in the coming year. The experimental results show that the LSTM model performs exceptionally well in predicting earthquakes of magnitude 5 < ML ≤ 6 within the time window of the test set, with a prediction success rate of 85%. Additionally, the study explores how different time windows, spatial locations, and parameter choices affect model performance. It found that longer time windows and key seismicity parameters, such as the b-value and the square root of total seismic energy, are crucial for improving prediction accuracy. Finally, we propose a magnitude interval-based assessment method to better predict the actual impacts that different magnitudes may cause. This method demonstrates the LSTM model’s potential in predicting moderate to strong earthquakes and offers new approaches for earthquake early warning and disaster mitigation.

Heart sound recognition using hybrid joint transform correlator

Abstract This work introduces an innovative method for heart sound recognition using hybrid joint transform correlator (JTC). The existing techniques such as convolutional neural networks and machine learning models, often suffer from computational complexity and latency difficulties, resulting in slower processing speeds. The proposed method addresses these challenges by utilizing an optical correlator, capable of operating at significantly accelerated speeds. The heart sound signals are first converted into spectrograms, which are then correlated using hybrid JTC operation for recognition and decision. To enhance the accuracy and robustness of the scheme, distortion-invariant filter fabrication has been incorporated with the JTC. The proposed approach not only enhances the speed of heart sound analysis but also maintains high precision and reliability. The effectiveness of the method has been confirmed through both simulation and experimental results.

Predicting responsiveness to GLP-1 pathway drugs using real-world data

BackgroundMedications targeting the glucagon-like peptide-1 (GLP-1) pathway are an important therapeutic class currently used for the treatment of Type 2 diabetes (T2D). However, there is not enough known about which subgroups of patients would receive the most benefit from these medications.ObjectiveThe goal of this study was to develop a predictive model for patient responsiveness to medications, here collectively called GLP-1 M, that include GLP-1 receptor agonists and dipeptidyl peptidase-4 (DPP4) inhibitors (that normally degrade endogenously-produced GLP-1). Such a model could guide clinicians to consider certain patient characteristics when prescribing second line medications for T2D.MethodsWe analyzed de-identified electronic health records of 7856 subjects with T2D treated with GLP-1 M drugs at Vanderbilt University Medical Center from 2003–2019. Using common clinical features (including commonly ordered lab tests, demographic information, other T2D medications, and diabetes-associated complications), we compared four different models: logistic regression, LightGBM, artificial neural network (ANN), and support vector classifier (SVC).ResultsOur analysis revealed that the traditional logistic regression model outperforms the other machine learning models, with an area under the Receiver Operating Characteristic curve (auROC) of 0.77.Our model showed that higher pre-treatment HbA1C is a dominant feature for predicting better response to GLP-1 M, while features such as use of thiazolidinediones or sulfonylureas is correlated with poorer response to GLP-1 M, as assessed by lowering of hemoglobin A1C (HbA1C), a standard marker of glycated hemoglobin used for assessing glycemic control in individuals with diabetes. Among female subjects under 40 taking GLP-1 M, the simultaneous use of non-steroidal anti-inflammatory drugs (NSAIDs) was associated with a greater reduction in HbA1C (0.82 ± 1.72% vs 0.28 ± 1.70%, p = 0.008).ConclusionThese findings indicate a thorough analysis of real-world electronic health records could reveal new information to improve treatment decisions for the treatment of T2D. The predictive model developed in this study highlights the importance of considering individual patient characteristics and medication interactions when prescribing GLP-1 M drugs.

Evaluating Chemical Transport and Machine Learning Models for Wildfire Smoke PM2.5: Implications for Assessment of Health Impacts.

Growing wildfire smoke represents a substantial threat to air quality and human health. However, the impact of wildfire smoke on human health remains imprecisely understood due to uncertainties in both the measurement of exposure of population to wildfire smoke and dose-response functions linking exposure to health. Here, we compare daily wildfire smoke-related surface fine particulate matter (PM2.5) concentrations estimated using three approaches, including two chemical transport models (CTMs): GEOS-Chem and the Community Multiscale Air Quality (CMAQ) and one machine learning (ML) model over the contiguous US in 2020, a historically active fire year. In the western US, compared against surface PM2.5 measurements from the US Environmental Protection Agency (EPA) and PurpleAir sensors, we find that CTMs overestimate PM2.5 concentrations during extreme smoke episodes by up to 3-5 fold, while ML estimates are largely consistent with surface measurements. However, in the eastern US, where smoke levels were much lower in 2020, CTMs show modestly better agreement with surface measurements. We develop a calibration framework that integrates CTM- and ML-based approaches to yield estimates of smoke PM2.5 concentrations that outperform individual approach. When combining the estimated smoke PM2.5 concentrations with county-level mortality rates, we find consistent effects of low-level smoke on mortality but large discrepancies in effects of high-level smoke exposure across different methods. Our research highlights the differences across estimation methods for understanding the health impacts of wildfire smoke and demonstrates the importance of bench-marking estimates with available surface measurements.

Exploring machine learning algorithms to predict not using modern family planning methods among reproductive age women in East Africa

BackgroundThe use of the modern family planning method provides chances for women to reach optimal child spacing, increase quality of life, increase economic status, achieve the desired family size, and prevent unsafe abortions and maternal and perinatal deaths. However, use modern family planning is low among reproductive age women in East Africa. Therefore, this study aimed at predicting not using modern family planning and identifying its associated factors among reproductive age women in east Africa using machine learning methods.MethodsThe data set was obtained from a demographic health survey in East Africa. This study used data on weighted sample of 92,564 women who were in their reproductive age. Python software was utilized for data processing, and machine learning models such as Random Forest (RF), Decision Tree (DT), XGB (Extreme Gradient Boosting), SVM (Support Victor Machine), and K-Nearest Neighbor Decision Tree (DT) were implemented. In this work, we evaluated the predictive models’ performance using performance assessment criteria such as accuracy, precision, recall, and the AUC curve.ResultIn this study, 92,564 reproductive women’s was used in the final analysis. Among the proposed machine learning models, XGB performed best overall in the proposed classifier, with an accuracy of 98.7%, precision of 99.8%, recall of 98%, 99.9% AUC score, had high specificity 98%, sensitivity 99.8%, and low error rate discovery was 0.013%. The most significant determinants of not using modern family planning in East Africa: lack of education, age 25–29, living in a rural area, being single, not knowing how to use a contraceptive, and smoking cigarette.ConclusionExtreme Gradient Boosting proved effective in predicting not using modern family planning, offering valuable insights for interventions and policy decisions in East Africa. Machine learning could help achieve early prediction and intervention on reproductive women’s not using modern family planning. This leads to a recommendation for policy direction to reduce child and maternal mortality and improve living standards in east Africa.

The role of gut microbiota and metabolomic pathways in modulating the efficacy of SSRIs for major depressive disorder

This study aims to explore the mechanism by which gut microbiota influences the antidepressant effects of serotonin reuptake inhibitors (SSRIs) through metabolic pathways. A total of 126 patients were analyzed for their gut microbiota and metabolomics. Patients received SSRI treatment and were categorized into responder and non-responder groups based on changes in their Hamilton Depression Rating Scale (HAMD-17) scores before and after treatment. The association between gut microbiota composition and the efficacy of SSRIs was investigated through 16S rRNA gene sequencing and metabolomic analysis, and a predictive model was developed. As a result, the study found significant differences in gut microbiota composition between the responder and resistant groups. Specific taxa, such as Ruminococcus, Bifidobacterium, and Faecalibacterium, were more abundant in the responder group. Functional analysis revealed upregulation of acetate degradation and neurotransmitter synthesis pathways in the responder group. The machine learning model indicated that gut microbiota and metabolites are potential biomarkers for predicting SSRIs efficacy. In conclusion, gut microbiota influences the antidepressant effects of SSRIs through metabolic pathways. The diversity and function of gut microbiota can serve as biomarkers for predicting the treatment response, providing new insights for personalized treatment.

Association and predictive ability between significant perioperative cardiovascular adverse events and stress glucose rise in patients undergoing non-cardiac surgery

BackgroundThe predictive importance of the stress hyperglycemia ratio (SHR), which is composed of admission blood glucose (ABG) and glycated hemoglobin (HbA1c), has not been fully established in noncardiac surgery. This study aims to evaluate the association and predictive capability the SHR for major perioperative adverse cardiovascular events (MACEs) in noncardiac surgery patients.MethodsIndividuals who underwent noncardiac surgical procedures between 2011 and 2020, including both diabetic and non-diabetic patients, were identified in the perioperative medicine database (INSPIRE 1.1) and classified into tertiles based on their SHR. The connection between the SHR and the risk of MACEs was studied using Cox proportional hazards regression analysis, then restricted cubic spline (RCS) was employed to assess the association’s form. Additionally, the SHR’s incremental predictive utility for MACEs was assessed by the C-statistic, continuous net reclassification improvement (NRI), and integrated discrimination improvement (IDI), thereby quantifying the enhancement in predictive accuracy brought by incorporating the SHR into existing risk models. Feature importance and predictive models were generated utilizing the Boruta algorithm and machine learning approaches.ResultsA total of 5609 patients were enrolled. With an upwards shift in SHR vertices, the rate of perioperative MACEs and cardiac death event steadily rose. The RCS analysis for perioperative MACEs and cardiac death event both indicated J-shaped associations. Inflection points occurred at SHR = 0.81 for MACEs and SHR = 0.97 for cardiac death. The model’s fit improved significantly, with a continuous NRI of 0.067 (95% CI: 0.025–0.137, P < 0.001) and an IDI of 0.305 (95% CI: 0.155–0.430, P < 0.001). When SHR was added as a categorical variable (> 0.81), the C-statistic increased to 0.785 (95% CI: 0.756–0.814) with a ΔC-statistic of 0.035 (P = 0.009), a continuous NRI of 0.007 (95% CI: 0.000-0.021, P = 0.016), and an IDI of 0.076 (95% CI -0.024-0.142, P = 0.092). In the Boruta algorithm, variables identified as important features in the green area were incorporated into the machine learning models development.ConclusionsThe SHR was related with an increased risk of perioperative MACEs in patients following noncardiac surgery, highlighting its potential as a useful and reliable predictive tool for assessing the risk of perioperative MACEs.Graphical

Urban Built Environment as a Predictor for Coronary Heart Disease—A Cross-Sectional Study Based on Machine Learning

The relationship between coronary heart disease (CHD) and complex urban built environments remains a subject of considerable uncertainty. The development of predictive models via machine learning to explore the underlying mechanisms of this association, as well as the formulation of intervention policies and planning strategies, has emerged as a pivotal area of research. A cross-sectional dataset of hospital admissions for CHD over the course of a year from a hospital in Dalian City, China, was assembled and matched with multi-source built environment data via residential addresses. This study evaluates five machine learning models, including decision tree (DT), random forest (RF), eXtreme gradient boosting (XGBoost), multi-layer perceptron (MLP), and support vector machine (SVM), and compares them with multiple linear regression models. The results show that DT, RF, and XGBoost exhibit superior predictive capabilities, with all R2 values exceeding 0.70. The DT model performed the best, with an R2 value of 0.818, and the best performance was based on metrics such as MAE and MSE. Additionally, using explainable AI techniques, this study reveals the contribution of different built environment factors to CHD and identifies the significant factors influencing CHD in cold regions, ranked as age, Digital Elevation Model (DEM), house price (HP), sky view factor (SVF), and interaction factors. Stratified analyses by age and gender show variations in the influencing factors for different groups: for those under 60 years old, Road Density is the most influential factor; for the 61–70 age group, house price is the top factor; for the 71–80 age group, age is the most significant factor; for those over 81 years old, building height is the leading factor; in males, GDP is the most influential factor; and in females, age is the most influential factor. This study explores the feasibility and performance of machine learning in predicting CHD risk in the built environment of cold regions and provides a comprehensive methodology and workflow for predicting cardiovascular disease risk based on refined neighborhood-level built environment factors, offering scientific support for the construction of sustainable healthy cities.

Predicting Rate Constants of Hydrogen Abstraction Reactions between OH/HO2 and Alkanes by Machine Learning Models.

The hydrogen abstraction reactions by small radicals from fuel molecules play an important role in the oxidation of fuels. However, experimental measurements and/or theoretical calculations of their rate constants under combustion conditions are very challenging due to their high reactivity. Machine learning offers a promising approach to predicting thermal rate constants. In this work, three machine learning methods, XGB, FNN, and XGB-FNN hybrid algorithms, were employed to train and predict the rate constants of the hydrogen abstraction reactions between alkanes and OH/HO2. Six descriptors were selected according to the Pearson correlation coefficients, the importance of descriptors, and the clustering heat map. It was proven that the XGB-FNN model is the most robust. The constructed XGB-FNN model achieved an average deviation of 89.13% for the alkanes + OH reactions and 190.93% for the alkanes + HO2 reactions on their respective prediction sets. The model was also used to predict the rate constants of the reactions involving larger alkanes, demonstrating its extrapolation capability. Furthermore, the model has the ability to distinguish the reactivity of the reactions with the hydrogen atom abstracted at different sites of alkane.

Profiling the AI speaker user: Machine learning insights into consumer adoption patterns.

The objective of this study is to identify the characteristics of users of AI speakers and predict potential consumers, with the aim of supporting effective advertising and marketing strategies in the fast-evolving media technology landscape. To do so, our analysis employs decision trees, random forests, support vector machines, artificial neural networks, and XGboost, which are typical machine learning techniques for classification and leverages the 2019 Media & Consumer Research survey data from the Korea Broadcasting and Advertising Corporation (N = 3,922). The final XGboost model, which performed the best among the other machine learning models, specifically forecasts individuals aged 45-50 and 60-65, who are active on social networking platforms and have a preference for varied programming content, as the most likely future users. Additionally, the model reveals their distinct lifestyle patterns, such as higher internet usage during weekdays and increased cable TV viewership on weekends, along with a better understanding of 5G technology. This pioneering effort in IoT consumer research employs advanced machine learning to not just predict, but intricately profile potential AI speaker consumers. It elucidates critical factors influencing technology uptake, including media consumption habits, attitudes, values, and leisure activities, providing valuable insights for creating focused and effective advertising and marketing strategies.

Incorporating Spatial Information for Regionalization of Environmental Parameters in Machine Learning Models

AbstractMachine learning models have gained popularity for environmental variable predictions due to their capacity to capture complex relationships and automate learning. However, incorporating spatial information as covariates into these models remains a challenge, as they may struggle to recognize spatial structures or autocorrelation without explicit training. In this study, we address this challenge by integrating spatial information into a random forest model, enhancing nitrate concentration predictions in groundwater. Using a dataset from 1,550 well locations in Baden-Wuerttemberg, Germany, spanning 2016 through 2019, we consider various environmental covariates including climate data, topography, land cover, soil properties, and hydrology. To incorporate spatial information, we employ eight techniques leveraging spatial coordinates (geographic coordinates, polynomial geographic coordinates, oblique geographic coordinates) or distances (Wendland transformed coordinates, Euclidean distance fields, Euclidean distance matrix, principal component analysis, eigenvector spatial filtering). Results are compared with a baseline model and a univariate ordinary kriging benchmark, evaluated through leave-one-out cross validation, various error metrics, and Moran’s I of residuals. Our findings highlight that integrating spatial information significantly enhances random forest model accuracy in predicting groundwater nitrate concentrations. Distance-based methods, like the Euclidean distance matrix, outperform coordinate-based approaches, albeit with higher computational requirements. Employing a dimension-reduced matrix strikes a balance between performance and accuracy. This study advances groundwater management and demonstrates the effectiveness of machine learning models in environmental studies.

Radiomics and deep learning features of pericoronary adipose tissue on non-contrast computerized tomography for predicting non-calcified plaques

Background Inflammation of coronary arterial plaque is considered a key factor in the development of coronary heart disease. Early the plaque detection and timely treatment of the atherosclerosis could effectively reduce the risk of cardiovascular events. However, there is no study combining radiomics with deep learning techniques to predict non-calcified plaques (NCP) in coronary artery at present. Objective To investigate the value of combination of radiomics and deep learning features based on non-contrast computerized tomography (CT) scans of pericoronary adipose tissue (PCAT), integrating with clinical risk factors of patients, in identifying coronary inflammation and predicting the presence of NCP. Methods The clinical and imaging data of 353 patients were analyzed. The region of interest (ROI) of PCAT was manually outlined on non-contrast CT scan images, like coronary CT calcium score sequential images, then the radiomics and deep learning features in ROIs were extracted respectively. In training set (Center 1), after performing feature selection, radiomics and deep learning feature models were established, meanwhile, clinical models were built. Finally, combined models were developed out via integrating clinical, radiomics, and deep learning features. The predictive performance of the four feature model groups (clinical, radiomics, deep learning, and three combination) was assessed by seven different machine learning models through generation of receiver operating characteristic curves (ROC) and the calculation of area under the curve (AUC), sensitivity, specificity, and accuracy. Furthermore, the predictive performance of each model was validated in an external validation set (Center 2). Results For the single model comparation, eXtreme Gradient Boosting (XGBoost) showed the best performance among the clinical model group in the validation set. And Random Forest (RF) exhibited the best indicative performance not only among the radiomics feature group but also in the deep learning feature model group. What's more, among the combined model group, RF still displayed the best predictive performance, with the value of AUC, sensitivity, specificity, and accuracy in the validation set are 0.963, 0.857, 0.929, and 0.905. Conclusion The RF model in the combined model group based on non-contrast CT scan PCAT can predict the presence of NCP more accurately and has the potential for preliminary screening of the NCP.

Detection and classification of long terminal repeat sequences in plant LTR-retrotransposons and their analysis using explainable machine learning

BackgroundLong terminal repeats (LTRs) represent important parts of LTR retrotransposons and retroviruses found in high copy numbers in a majority of eukaryotic genomes. LTRs contain regulatory sequences essential for the life cycle of the retrotransposon. Previous experimental and sequence studies have provided only limited information about LTR structure and composition, mostly from model systems. To enhance our understanding of these key sequence modules, we focused on the contrasts between LTRs of various retrotransposon families and other genomic regions. Furthermore, this approach can be utilized for the classification and prediction of LTRs.ResultsWe used machine learning methods suitable for DNA sequence classification and applied them to a large dataset of plant LTR retrotransposon sequences. We trained three machine learning models using (i) traditional model ensembles (Gradient Boosting), (ii) hybrid convolutional/long and short memory network models, and (iii) a DNA pre-trained transformer-based model using k-mer sequence representation. All three approaches were successful in classifying and isolating LTRs in this data, as well as providing valuable insights into LTR sequence composition. The best classification (expressed as F1 score) achieved for LTR detection was 0.85 using the hybrid network model. The most accurate classification task was superfamily classification (F1=0.89) while the least accurate was family classification (F1=0.74). The trained models were subjected to explainability analysis. Positional analysis identified a mixture of interesting features, many of which had a preferred absolute position within the LTR and/or were biologically relevant, such as a centrally positioned TATA-box regulatory sequence, and TG..CA nucleotide patterns around both LTR edges.ConclusionsOur results show that the models used here recognized biologically relevant motifs, such as core promoter elements in the LTR detection task, and a development and stress-related subclass of transcription factor binding sites in the family classification task. Explainability analysis also highlighted the importance of 5’- and 3’- edges in LTR identity and revealed need to analyze more than just dinucleotides at these ends. Our work shows the applicability of machine learning models to regulatory sequence analysis and classification, and demonstrates the important role of the identified motifs in LTR detection.

Uncovering the Diagnostic Power of Radiomic Feature Significance in Automated Lung Cancer Detection: An Integrative Analysis of Texture, Shape, and Intensity Contributions

Background: Lung cancer still maintains the leading position among causes of death in the world; the process of early detection surely contributes to changes in the survival of patients. Standard diagnostic methods are grossly insensitive, especially in the early stages. In this paper, radiomic features are discussed that can assure improved diagnostic accuracy through automated lung cancer detection by considering the important feature categories, such as texture, shape, and intensity, originating from the CT DICOM images. Methods: We developed and compared the performance of two machine learning models—DenseNet-201 CNN and XGBoost—trained on radiomic features with the ability to identify malignant tumors from benign ones. Feature importance was analyzed using SHAP and techniques of permutation importance that enhance both the global and case-specific interpretability of the models. Results: A few features that reflect tumor heterogeneity and morphology include GLCM Entropy, shape compactness, and surface-area-to-volume ratio. These performed excellently in diagnosis, with DenseNet-201 producing an accuracy of 92.4% and XGBoost at 89.7%. The analysis of feature interpretability ascertains its potential in early detection and boosting diagnostic confidence. Conclusions: The current work identifies the most important radiomic features and quantifies their diagnostic significance through a properly conducted feature selection process reflecting stability analysis. This provides the blueprint for feature-driven model interpretability in clinical applications. Radiomics features have great value in the automated diagnosis of lung cancer, especially when combined with machine learning models. This might improve early detection and open personalized diagnostic strategies for precision oncology.

Exploring Generative Adversarial Network-Based Augmentation of Magnetic Resonance Brain Tumor Images

Background: A generative adversarial network (GAN) has gained popularity as a data augmentation technique in the medical field due to its efficiency in creating synthetic data for different machine learning models. In particular, the earlier literature suggests that the classification accuracy of a convolutional neural network (CNN) used for detecting brain tumors in magnetic resonance imaging (MRI) images increases when GAN-generated images are included in the training data together with the original images. However, there is little research about how the exact number of GAN-generated images and their ratio to the original images affects the results obtained. Materials and methods: Here, by using 1000 original images from a public repository with MRI images of patients with or without brain tumors, we built a GAN model to create synthetic brain MRI images. A modified U-Net CNN is trained multiple times with different training datasets and its classification accuracy is evaluated from a separate test set of another 1000 images. The Mann–Whitney U test is used to estimate whether the differences in the accuracy caused by different choices of training data are statistically significant. Results: According to our results, the use of GAN augmentation only sometimes produces a significant improvement. For instance, the classification accuracy significantly increases when 250–750 GAN-generated images are added to 1000 original images (p-values ≤ 0.0025) but decreases when 10 GAN-generated images are added to 500 original images (p-value: 0.03). Conclusions: Whenever GAN-based augmentation is used, the number of GAN-generated images should be carefully considered while accounting for the number of original images.

Predicting maternal risk level using machine learning models

BackgroundMaternal morbidity and mortality remain critical health concerns globally. As a result, reducing the maternal mortality ratio (MMR) is part of goal 3 in the global sustainable development goals (SDGs), and previously, it was an important indicator in the Millennium Development Goals (MDGs). Therefore, identifying high-risk groups during pregnancy is crucial for decision-makers and medical practitioners to mitigate mortality and morbidity. However, the availability of accurate predictive models for maternal mortality and maternal health risks is challenging. Compared with traditional predictive models, machine learning algorithms have emerged as promising predictive modelling methods providing accurate predictive models.MethodsThis work aims to explore the potential of machine learning (ML) algorithms in maternal risk level prediction using a nationwide maternal mortality dataset from Oman for the first time. A total of 402 maternal deaths from 1991 to 2023 in Oman were included in this study. We utilised principal component analysis (PCA) in the ML algorithms and compared them to the results of model performance without PCA. We employed and compared ten ML algorithms, including decision tree (DT), random forest (RF), K—Nearest Neighbors (KNN), Naïve Bayes (NB), Extreme Gradient Boosting (xgboost), Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Logistic Regression (LR), Support Vector Machine (SVM) and Artificial Neural Network (ANN). Different metrics, including, accuracy, sensitivity, precision, and the F1- score, were utilised to assess Model performance.ResultsThe results indicated that the RF model outperformed the other methods in predicting the risk level (low or high) with an accuracy of 75.2%, precision of 85.7% and F1- score of 73% after PCA was applied.ConclusionsWe applied several machine learning models to predict maternal risk levels for the first time using real data from Oman. RF outperformed the other algorithms in this classification problem. A reliable estimate of maternal risk level would facilitate intervention plans for medical practitioners to reduce maternal death.

Auxiliary identification of depression patients using interpretable machine learning models based on heart rate variability: a retrospective study

ObjectiveDepression has emerged as a global public health concern with high incidence and disability rates, which are timely imperative to identify and intervene in clinical practice. The objective of this study was to explore the association between heart rate variability (HRV) and depression, with the aim of establishing and validating machine learning models for the auxiliary diagnosis of depression.MethodsThe data of 465 outpatients from the Affiliated Hospital of Southwest Medical University were selected for the study. The study population was then randomly divided into training and test sets in a 7:3 ratio. Logistic regression (LR), support vector machine (SVM), random forest (RF) and eXtreme gradient boosting (XGBoost) algorithm models were used to construct risk prediction models in the training set, and the model performance was verified in the test set. The four models were evaluated by the area under the receiver operating characteristic curve (ROC), calibration curve and the decision curve analysis (DCA). Furthermore, we employed the SHapley Additive exPlanations (SHAP) method to illustrate the effects of the features attributed to the model.ResultsThere were 237 people in the depressed group and 228 in the non-depressed group. In the training set (n = 325) and test set (n = 140), the area under of the curve(AUC) values of the XGBoost model are 0.92 [95% confidence interval (CI) 0.888,0.95] and 0.82 (95% CI 0.754,0.892)] respectively, which are higher than the other three models. The XGBoost model has excellent predictive efficacy and clinical utility. The SHAP method was ranked according to the importance of the degree of influence on the model, with age, heart rate, Standard deviation of the NN intervals (SDNN), two nonlinear parameters of HRV and sex considered to be the top 6 predictors.ConclusionWe provided a feasibility study of HRV as a potential biomarker for depression. The proposed model based on HRV provides clinicians with a quantitative auxiliary diagnostic tool, which is assist to improving the accuracy and efficiency of depression diagnosis, and can also be utilized for the monitoring and prevention of depression.

An interpretable machine learning model for precise prediction of biomarkers for intermittent fasting pattern

Intermittent fasting is currently a highly sought-after dietary pattern. To explore the potential biomarkers of intermittent fasting, untargeted metabolomics analysis of fecal metabolites in two groups of mice, intermittent fasting and normal feeding, was conducted using UPLC-HRMS. The data was further analyzed through interpretable machine learning (ML) to data mine the biomarkers for two dietary patterns. We developed five machine learning models and results showed that under three-fold cross-validation, Random Forest model was the most suitable for distinguishing the two dietary patterns. Finally, Shapely Additive exPlanations (SHAP) were explored to perform a weighted explanatory analysis on the Random Forest model, and the contribution of each metabolite to the model was calculated. Results indicated that Ganoderenic Acid C is the potential biomarkers to distinguish the two dietary patterns. Our work provides new insights for metabolic biomarker analysis and lays a theoretical foundation for the selection of a healthieir dietary lifestyle.