Machine learning algorithms for outcome prediction in(chemo)radiotherapy: An empirical comparison of classifiers

ObjectiveThis study aims to develop and compare different models to predict the Length of Stay (LoS) and the Prolonged Length of Stay (PLoS) of inpatients admitted through the emergency department (ED) in general patient settings. This aim is not only to promote any specific model but rather to suggest a decision-supporting tool (i.e., a prediction framework).MethodsWe analyzed a dataset of patients admitted through the ED to the “Sant”Orsola Malpighi University Hospital of Bologna, Italy, between January 1 and October 26, 2022. PLoS was defined as any hospitalization with LoS longer than 6 days. We deployed six classification algorithms for predicting PLoS: Random Forest (RF), Support Vector Machines (SVM), Gradient Boosting (GB), AdaBoost, K-Nearest Neighbors (KNN), and logistic regression (LoR). We evaluated the performance of these models with the Brier score, the area under the ROC curve (AUC), accuracy, sensitivity (recall), specificity, precision, and F1-score. We further developed eight regression models for LoS prediction: Linear Regression (LR), including the penalized linear models Least Absolute Shrinkage and Selection Operator (LASSO), Ridge and Elastic-net regression, Support vector regression, RF regression, KNN, and eXtreme Gradient Boosting (XGBoost) regression. The model performances were measured by their mean square error, mean absolute error, and mean relative error. The dataset was randomly split into a training set (70%) and a validation set (30%).ResultsA total of 12,858 eligible patients were included in our study, of whom 60.88% had a PloS. The GB classifier best predicted PloS (accuracy 75%, AUC 75.4%, Brier score 0.181), followed by LoR classifier (accuracy 75%, AUC 75.2%, Brier score 0.182). These models also showed to be adequately calibrated. Ridge and XGBoost regressions best predicted LoS, with the smallest total prediction error. The overall prediction error is between 6 and 7 days, meaning there is a 6–7 day mean difference between actual and predicted LoS.ConclusionOur results demonstrate the potential of machine learning-based methods to predict LoS and provide valuable insights into the risks behind prolonged hospitalizations. In addition to physicians' clinical expertise, the results of these models can be utilized as input to make informed decisions, such as predicting hospitalizations and enhancing the overall performance of a public healthcare system.

P125. Development of a novel ensemble machine learning algorithm for prediction of complications and readmission after anterior cervical spinal fusion

This study aims to determine the feasibility of machine learning (ML) and patient registration record to be utilised to develop an over-the-counter (OTC) screening model for breast cancer risk estimation. Data were retrospectively collected from women who came to the Hospital Universiti Sains Malaysia, Malaysia for breast-related problems. Eight ML models were used: k-nearest neighbour (kNN), elastic-net logistic regression, multivariate adaptive regression splines, artificial neural network, partial least square, random forest, support vector machine (SVM), and extreme gradient boosting. Features utilised for the development of the screening models were limited to information in the patient registration form. The final model was evaluated in terms of performance across a mammographic density. Additionally, the feature importance of the final model was assessed using the model agnostic approach. kNN had the highest Youden J index, precision, and PR-AUC, while SVM had the highest F2 score. The kNN model was selected as the final model. The model had a balanced performance in terms of sensitivity, specificity, and PR-AUC across the mammographic density groups. The most important feature was the age at examination. In conclusion, this study showed that ML and patient registration information are feasible to be used as the OTC screening model for breast cancer.

Category:Ankle Arthritis; AnkleIntroduction/Purpose:Ankle arthrodesis and total ankle replacement are the most commonly performed procedures for surgical management of ankle arthritis. Arthrodesis provides effective pain relief but the rate of complications after arthrodesis is higher as it is more commonly performed in patients with comorbidities that preclude ankle replacement. Accurately risk- stratifying patients who undergo ankle arthrodesis would be of great utility, given the significant cost and morbidity associated with developing major perioperative complications. There is a paucity of accurate prediction models that can be used to pre- operatively risk-stratify patients for ankle arthrodesis. We aim to develop a machine learning (ML) algorithm for prediction of major perioperative complication after ankle arthrodesis as well as compare its performance against traditional predictive models based on logistic regression.Methods:This is a retrospective cohort study of adult patients who underwent ankle arthrodesis at any non-federal California hospital between 2015 and 2017. The primary outcome was readmission within 30 days or major perioperative complication - venous thromboembolism within 30 days, myocardial infarction within 7 days, pneumonia within 7 days, systemic infection within 7 days, surgical site bleeding within 90 days, and wound complications within 90 days. We build ML and logistic regression models that span different classes of modeling approaches: XGBoost, AdaBoost, Gradient Boosting, and Random Forest. Discrimination and calibration were assessed using area under the receiver operating characteristic curve (AUROC) and Brier score, respectively. We utilize a partial dependence function to measure the importance of an individual feature by assessing the average effect in predicted risks when its value is altered. We rank the contribution of the included variables to the prediction of adverse outcomes.Results:A total of 1,084 patients met inclusion criteria for this study. There were 131 major complications or readmission (12.1%). The optimized XGBoost algorithm demonstrates higher discrimination (AUROC: 0.707 + 0.052) compared to LR (0.691 + 0.055). The receiver operating characteristic curves for the XGBoost and logistic regression models are visualized in Figure 1. XGBoost also outperforms the three other ML models. This model was well calibrated (Brier score: 0.103 + 0.001). The variables most important for the XGBoost model include diabetes, chronic kidney disease, implant complication, and major fracture. Five of the ten most important features for XGBoost were markedly less important for the traditional logistic regression model: male sex, prior hip fracture, cardiorespiratory failure, acute renal failure, and dialysis status.Conclusion:We report a ML algorithm for prediction of major perioperative complications after ankle arthrodesis. The optimized XGBoost model is well-calibrated and demonstrates superior risk prediction to logistic regression. This tool may identify and address potentially modifiable risk factors, helping to accurately risk-stratify patients and decrease likelihood of major complications. Notably, the predictors most important for XGBoost are different from those for logistic regression. This suggests that the superior discriminative capability of ML methods stems from their ability to capture complex non-linear relationships between variables that logistic regression is unable to detect.

Accurate forecasting of photovoltaic (PV) generated electricity is essential for efficiently managing and integrating Renewable Energy (RE) into electricity distribution systems. This research investigation optimizes Feature Selection (FS) and prediction results for PV energy prediction by applying Bayesian Density Estimation (BDE) with Elastic Net (ELNET) regression analysis. This phenomenon and unacceptable outcomes are prevalent when applying conventional regression algorithms on datasets with significant results and addressing predictor multicollinearity. Improved FS and multicollinearity control has been rendered feasible by ELNET, which integrates the best features of Ridge and Lasso regression. ELNET eliminates these challenges through the implementation of L1 and L2 penalties. Non-parametric prediction Bayesian Density Estimation (BDE) is comprehensive data regarding residual distributions and predictor impacts. By incorporating ELNET’s regularisation and FS abilities with BDE’s statistical prediction and adaptability, the recommended ELNET-BDE is proposed to attain more accurate and reliable predictions. This technique has been used to assess massive data sets developing from Visakhapatnam, India, incorporating historical PV energy generation combined with definite Meteorological Factors (MF). Considering comprehensive data preliminary processing, FS, and validation, ELNET-BDE outperforms existing methods. Research investigations demonstrate that the ELNET-BDE model attains significantly lower Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) than contesting Machine Learning (ML) algorithms like Artificial Neural Network (ANN), Support Vector Machine (SVM), Random Forest (RF), and Gradient Boosting Machines (GBM). Compared with distinct FS techniques, RMSE can be minimized by up to 15% and MAE by up to 20%. The findings specify a substantial improvement in accuracy in prediction, emphasizing how the model can be used for improving solar power grid integration and energy for improved RE management.

Background Malaria remains a major cause of illness and death among children under five in Nigeria, despite efforts to control transmission. Accurate and reliable prediction of malaria outbreaks is crucial for health authorities to take timely measures. This study aims to identify the most robust machine learning classification algorithms for predicting the status of malaria in children under five (0-59 months). Methods The 2021 Nigeria Malaria Indicator Survey (NMIS) included 10,655 children under five who were tested for malaria using the Rapid Diagnostic Test (RDT). Various machine learning models were explored, including Decision Trees, K-Nearest Neighbor, Naïve Bayes, Random Forest, Support Vector Machines, and Survey Logistic Regression, and their performance was evaluated through metrics such as accuracy, AUC, balanced accuracy, F1-Score, negative predictive value, precision, sensitivity, and specificity. Results Random Forest (RF) is the most robust and balanced classification model due to its superior accuracy (79%), precision (77%), recall (62%), F1-score (69%), and AUC (80%). Support Vector Machine (SVM) also demonstrated strong performance, particularly in accuracy (74%) and AUC (80%). Survey Logistic Regression (SLR) and Decision Tree (DT) offered moderate results but fell short compared to RF and SVM, indicating the need for further optimization. Naive Bayes (NB) and K-Nearest Neighbors (KNN) had limitations, making them less reliable for this task. Conclusion In conclusion, the study reveals that RF and SVM are the best classification models for predicting malaria status in children under five years old. RF is reliable and balanced, while SVM is preferred for recall. SLR and DT show potential but require optimization. NB and KNN have significant performance gaps, making them less suitable. These findings will help policymakers and malaria intervention programs address key factors, enabling more targeted public health interventions to reduce the malaria burden on young children and improve the well-being of vulnerable populations in Nigeria.

Development of supervised machine learning algorithms for prediction of satisfaction at 2 years following total shoulder arthroplasty

BackgroundMachine learning (ML) has shown great potential in recognizing complex disease patterns and supporting clinical decision-making. Diabetic foot ulcers (DFUs) represent a significant multifactorial medical problem with high incidence and severe outcomes, providing an ideal example for a comprehensive framework that encompasses all essential steps for implementing ML in a clinically relevant fashion.ObjectiveThis paper aims to provide a framework for the proper use of ML algorithms to predict clinical outcomes of multifactorial diseases and their treatments.MethodsThe comparison of ML models was performed on a DFU dataset. The selection of patient characteristics associated with wound healing was based on outcomes of statistical tests, that is, ANOVA and chi-square test, and validated on expert recommendations. Imputation and balancing of patient records were performed with MIDAS (Multiple Imputation with Denoising Autoencoders) Touch and adaptive synthetic sampling, respectively. Logistic regression, support vector machine (SVM), k-nearest neighbors, random forest (RF), extreme gradient boosting (XGBoost), Bayesian additive regression trees, and artificial neural network were trained, cross-validated, and optimized using random sampling on the patient dataset. To evaluate model calibration and clinical utility, calibration curves, Brier scores, and decision curve analysis (DCA) were performed.ResultsThe exploratory dataset consisted of 700 patient records with 199 variables. After dataset cleaning, the variables used for model training included age, smoking status, toe systolic pressure, blood pressure, oxygen saturation, hemoglobin, hemoglobin A1c, estimated glomerular filtration rate, wound location, diabetes type, Texas wound classification, neuropathy, and wound area measurement. The SVM obtained a stable accuracy of 0.853 (95% CI 0.810-0.896) with an area under the receiver operating characteristic curve of 0.922 (95% CI 0.889-0.955). The RF and XGBoost acquired an accuracy of 0.838 (95% CI 0.793-0.883) and 0.815 (95% CI 0.768-0.862), respectively, with areas under the receiver operating characteristic curve of 0.917 (95% CI 0.883-0.951) for RF and 0.889 (95% CI 0.849-0.929) for XGBoost. SVM, RF, and XGBoost were well-calibrated, with average Brier scores around 0.127 (SD 0.013). DCA showed that the SVM provided the highest net clinical benefit across relevant risk thresholds.ConclusionsHandling missing values, feature selection, and addressing class imbalance are critical components of the key steps in developing ML applications for clinical research. Seven models were selected for comparing their predictive power regarding complete wound healing, and each model representing a different branch in ML. In this initial DFU dataset used as an example, the SVM achieved the best performance in predicting clinical outcomes, followed by RF and XGBoost. The model’s calibration and clinical utility were determined through calibration curves, Brier scores, and DCA, demonstrating its potential relevance in clinical decision-making.

Alzheimer’s disease (AD) is a gradient degeneration of essential cognitive activities such as memory, thinking, and cognition. AD mainly affects elderly individuals and is recognized as the most common cause of dementia. This study investigates the predictive performance of nine supervised machine learning algorithms—Logistic Regression, Decision Tree, Random Forest, K-Nearest Neighbors, Support Vector Machine, Gaussian Naïve Bayes, Multi-Layer Perceptron, eXtreme Gradient Boost, and Gradient Boosting—using neuropsychological assessment data. We applied two classification techniques—binary and multiclass—to classify 1761 subjects into three categories: cognitively normal (CN), mild cognitive impairment (MCI), and Alzheimer's disease (AD). Binary classification tasks focused on CNvsAD and CNvsMCI subsets, while multiclass classification used the full dataset (TriClass). Hyperparameter tuning was performed to optimize model performance. The results indicate that ensemble learning models, particularly Gradient Boosting (GB) and Random Forest (RF), exhibited superior accuracy compared to other algorithms. Most models for the CNvsAD subset achieved the highest accuracy (97.74%), while GB achieved the best performance (94.98%) for the CNvsMCI subset. For multiclass classification, RF achieved the highest accuracy at 84.70%. These findings highlight the robustness and efficiency of ensemble learning algorithms, especially in handling complex, non-linear data structures. This study underscores the potential of RF and GB as reliable tools for early detection and classification of Alzheimer’s disease using neuropsychological data.

647 Background: The primary treatment of most mNSTC is chemotherapy followed by surgery if the residual disease (RD) is >1 cm. However, conventional imaging lacks the specificity to characterize the tissue, often leading to overtreatment. This study hypothesizes that integrating CT-driven radiomics features with plasma miR371 and miR375 will enhance the predictive accuracy of Machine Learning (ML) models to predict teratoma, viable germ cell (vGCT) and fibrosis/necrosis (F/N) in mNSTC patients with RD. Methods: 111 lesions from52 patients, including residual teratoma (n=57), F/N (n=33), vGCT (n=10), and additional seminoma (n=11) for training purposes were included, split into training (N=78) and test cohorts (N=33). Lesions were lymph nodes (n=87), lung (n=21), and brain (n=3) with a median size of 1.6 cm (Q1-Q3 interval=1.2-2.73 cm). 3D Slicer version 5.6.1 was used to segment the RD > 1 cm (short axis) and extract radiomics features. Plasma miRNA levels before resection were measured by RT-PCR. Random Forest (RF), Support Vector Machine (SVM), Gradient Boosting (GB), and CatBoost (CB) ML models were evaluated to define the operating characteristics of radiomics alone (R-only) and in combination with miR371 (371) and/or miR375 (375) levels in predicting teratoma, vGCT and F/N. Results: For predicting teratoma, the best models were RF (R+375 and R+371+375), CB (R+371+375), and GB (R+371 and R+371+375). While adding miR371 or miR375 to R-only slightly improved AUC across models, the best results were achieved with the R+375+371 dataset. CB achieved AUCs ranging from 0.94 to 0.97 in training and 0.81 to 0.93 in test sets, with its highest AUC of 0.93 (95% CI: 0.78-0.97) on the R+375+371 dataset to differentiate all three classes. Similarly, GB demonstrated strong performance, achieving its highest AUC of 0.93 (95% CI: 0.79-0.96) on the R+375+371 dataset (Table). Conclusions: Integration of plasma miR371, miR375 and radiomics improved accuracy of predicting histologies across all ML models. These methods could be used to characterize the histology of RD in mNSTC patients to better inform treatment decisions. Further refinement, including incorporation of histological findings of the primary tumor, will be reported. AUC values of different ML algorithms on training and test sets. TRAINING SET TEST SET Model ±SD R R+375 R+371 R+375+371 Model (95% CI) R R+375 R+371 R+375+371 RF 0.93±0.05 0.95±0.04 0.95±0.03 0.96±0.04 RF 0.8(0.59-0.89) 0.85(0.72-0.93) 0.87(0.76-0.95) 0.91(0.78-0.95) SVM 0.84±0.06 0.84±0.09 0.89±0.11 0.89±0.09 SVM 0.72(0.54-0.80) 0.74(0.56-0.82) 0.83(0.69-0.92) 0.84(0.76-0.94) GB 0.94±0.04 0.91±0.08 0.95±0.05 0.97±0.03 GB 0.84(0.61-0.96) 0.89(0.77-0.97) 0.89(0.79-0.96) 0.93(0.79-0.96) CB 0.95±0.03 0.94±0.03 0.94±0.04 0.97±0.03 CB 0.81(0.6-0.93) 0.86(0.73-0.94) 0.89(0.78-0.97) 0.93(0.78-0.97)

The prevalence of urinary tract infections (UTIs) is a significant concern, particularly among women. This study aims to adopt machine learning models for predicting UTIs using gradient boosting, support vector machine (SVM), and random forest algorithms. The dataset used was obtained from the Kaggle repository, consisting of UTI articles. The performance of the models was evaluated using metrics such as accuracy, precision, recall, and F1-score. The UTI dataset was preprocessed by removing irrelevant columns and handling missing values. The dataset was then divided into training and testing sets. The three models were trained using the training data. The results showed that both random forest and gradient boosting classification achieved an accuracy of 0.8750, while SVM achieved an accuracy of 0.8333. Precision values were 0.8000 for random forest, 1.0000 for SVM, and 0.9000 for gradient boosting. F1-scores were 0.8421 for both random forest and SVM, and 0.8571 for gradient boosting. Recall values were 0.8889 for both random forest and SVM, and 0.8182 for gradient boosting. Comparing the results with previous studies, the models showed competitive performance. The ROC curves were plotted to visualize the models' performance across different classification thresholds. In conclusion, the study demonstrates the effectiveness of gradient boosting, SVM, and random forest algorithms in predicting UTIs. The adopted models provide valuable insights for early detection and management of UTIs, contributing to improved healthcare outcomes.

The prevalence of nitrates (NO3−) in groundwater due to the extensive application of fertilizers and anthropogenic sources pollutes the groundwater. Machine learning (ML) techniques are now being increasingly deployed to achieve high precision in predicting water quality. This study assesses the efficacy of nine distinct ML algorithms, namely, linear regression, polynomial regression, decision tree, random forest (RF), support vector machine, multilayer perceptron regressor, eXtreme gradient boosting (XGB), light gradient boosting (LGB), and K‐nearest neighbors to predict nitrate concentration in the groundwater in Thiruvannamalai District, Tamil Nadu. Overall, 360 water samples for 1 year and 14 groundwater variables were determined to predict nitrate. Performance evaluation metrics such as root mean square error (RMSE), moving average error (MAE), and correlation coefficient (R2) were evaluated for pre‐monsoon, monsoon, and post‐monsoon seasons. For all three seasons, RF predicted the nitrate concentration with low values of RMSE, MAE, and higher values of R2. The results show values for RF with: RSME: 0.49, MAE: 1.30, and R2: 0.94, which has a higher prediction tailed by LGB and XGB and is true for all the seasons. The results from the study will aid the policymakers in planning the strategy for remediation.

ABSTRACTThe prediction of pollutants removal efficiency from the generated effluent of a treatment plant is valuable and can reduce the time, sampling and energy required during performance assessment. The present study aims to predict the effect of different input parameters on the treatment efficiency of the developed microbial‐based anaerobic process for textile effluent using machine leaning algorithms. The decolourisation and chemical oxygen demand (COD) reduction of the treated effluent were predicted on the basis of the three different input parameters pH, COD and colour value of the textile wastewater. The effectiveness of different machine learning algorithms, support vector machines (SVM), random forest (RF), gradient boost regressor (GBR), AdaBoost, extreme gradient boosting (XGB) regressor and voting regressor, were evaluated based on the correlation coefficient (R2) value. The results revealed that the RF achieved the highest accuracy for decolourisation (training data R2: ∼0.85 and test data R2: ∼0.84) as well as COD reduction (training data R2: ∼0.87 and test data R2: ∼0.94) compared to the other algorithms. These results were validated experimentally, confirming that RF can be used as a tool to predict the performance efficiency of a microbial‐based treatment system.

Hyperparameter tuning and performance assessment of statistical and machine-learning algorithms using spatial data

Hepatitis B virus (HBV) infection remains a public health concern in China, especially among middle-aged and older adults. Accurate prediction models can assist in early detection and targeted interventions. However, predicting HBV infection using population-level data is challenging due to its relatively low prevalence, which creates highly imbalanced datasets that undermine the performance of traditional predictive models. This study aimed to develop a machine learning model to predict HBV infection using an ensemble approach designed to handle class imbalance. Our study conducted a cross-sectional analysis of participants aged 45 and above from the China Health and Retirement Longitudinal Study 2011. To address the class imbalance in HBV infection, the synthetic minority oversampling technique (SMOTE) was employed. We developed machine learning models, including logistic regression (LR), support vector machine (SVM), random forest (RF), gradient boosting machine (GBM), adaptive boosting classifier (AdaBoost), extreme gradient boosting (XGBoost), and stacking ensemble model (SEM), to predict HBV infection and identify predictors. Model performance was evaluated using area under the receiver operating characteristic curve (AUC), area under the precision-recall curve (AP), accuracy, sensitivity, specificity, precision, F1 score, and Brier score. Shapley additive explanation (SHAP) analysis was applied to interpret each predictor’s individual contribution and direction to the model’s output. In the final analysis, 8,954 participants were included, among whom 0.64% self-reported HBV infection, confirming highly imbalanced data. The SEM model achieved balanced performance (AUC: 87.93%, AP: 18.85%, F1: 15.66%, Brier score: 0.03). Machine learning models also identified several new important predictors, including birth season, depression, and knee height. An ensemble machine learning model incorporating techniques to manage data imbalance can accurately predict HBV infection in Chinese middle-aged and older adults. The novel predictors we identified offer new insights into HBV risk in ageing populations and support the development of more targeted prevention strategies.