Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Purpose: Biomarkers predicting clinically significant prostate cancer (sPC) before biopsy are currently lacking. This study aimed to develop a non-invasive urine test to predict sPC in at-risk men using urinary metabolomic profiles. Materials and Methods: Urine samples from 934 at-risk subjects and 268 treatment-naïve PC patients were subjected to liquid chromatography/mass spectrophotometry (LC-MS)-based metabolomics profiling using both C18 and hydrophilic interaction liquid chromatography (HILIC) column analyses. Four models were constructed (training cohort [n=647]) and validated (validation cohort [n=344]) for different purposes. Model I differentiates PC from benign cases. Models II, III, and a Gleason score model (model GS) predict sPC that is defined as National Comprehensive Cancer Network (NCCN)-categorized favorable-intermediate risk group or higher (Model II), unfavorable-intermediate risk group or higher (Model III), and GS ≥7 PC (model GS), respectively. The metabolomic panels and predicting models were constructed using logistic regression and Akaike information criterion. Results: The best metabolomic panels from the HILIC column include 25, 27, 28 and 26 metabolites in Models I, II, III, and GS, respectively, with area under the curve (AUC) values ranging between 0.82 and 0.91 in the training cohort and between 0.77 and 0.86 in the validation cohort. The combination of the metabolomic panels and five baseline clinical factors that include serum prostate-specific antigen, age, family history of PC, previously negative biopsy, and abnormal digital rectal examination results significantly increased AUCs (range 0.88–0.91). At 90% sensitivity (validation cohort), 33%, 34%, 41%, and 36% of unnecessary biopsies were avoided in Models I, II, III, and GS, respectively. The above results were successfully validated using LC-MS with the C18 column. Conclusions Urinary metabolomic profiles with baseline clinical factors may accurately predict sPC in men with elevated risk before biopsy.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Purpose: Biomarkers predicting clinically significant prostate cancer (sPC) before biopsy are currently lacking. This study aimed to develop a non-invasive urine test to predict sPC in at-risk men using urinary metabolomic profiles. Materials and Methods: Urine samples from 934 at-risk subjects and 268 treatment-naïve PC patients were subjected to liquid chromatography/mass spectrophotometry (LC-MS)-based metabolomics profiling using both C18 and hydrophilic interaction liquid chromatography (HILIC) column analyses. Four models were constructed (training cohort [n=647]) and validated (validation cohort [n=344]) for different purposes. Model I differentiates PC from benign cases. Models II, III, and a Gleason score model (model GS) predict sPC that is defined as National Comprehensive Cancer Network (NCCN)-categorized favorable-intermediate risk group or higher (Model II), unfavorable-intermediate risk group or higher (Model III), and GS ≥7 PC (model GS), respectively. The metabolomic panels and predicting models were constructed using logistic regression and Akaike information criterion. Results: The best metabolomic panels from the HILIC column include 25, 27, 28 and 26 metabolites in Models I, II, III, and GS, respectively, with area under the curve (AUC) values ranging between 0.82 and 0.91 in the training cohort and between 0.77 and 0.86 in the validation cohort. The combination of the metabolomic panels and five baseline clinical factors that include serum prostate-specific antigen, age, family history of PC, previously negative biopsy, and abnormal digital rectal examination results significantly increased AUCs (range 0.88–0.91). At 90% sensitivity (validation cohort), 33%, 34%, 41%, and 36% of unnecessary biopsies were avoided in Models I, II, III, and GS, respectively. The above results were successfully validated using LC-MS with the C18 column. Conclusions Urinary metabolomic profiles with baseline clinical factors may accurately predict sPC in men with elevated risk before biopsy.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Purpose: Biomarkers predicting clinically significant prostate cancer (sPC) before biopsy are currently lacking. This study aimed to develop a non-invasive urine test to predict sPC in at-risk men using urinary metabolomic profiles. Materials and Methods: Urine samples from 934 at-risk subjects and 268 treatment-naïve PC patients were subjected to liquid chromatography/mass spectrophotometry (LC-MS)-based metabolomics profiling using both C18 and hydrophilic interaction liquid chromatography (HILIC) column analyses. Four models were constructed (training cohort [n=647]) and validated (validation cohort [n=344]) for different purposes. Model I differentiates PC from benign cases. Models II, III, and a Gleason score model (model GS) predict sPC that is defined as National Comprehensive Cancer Network (NCCN)-categorized favorable-intermediate risk group or higher (Model II), unfavorable-intermediate risk group or higher (Model III), and GS ≥7 PC (model GS), respectively. The metabolomic panels and predicting models were constructed using logistic regression and Akaike information criterion. Results: The best metabolomic panels from the HILIC column include 25, 27, 28 and 26 metabolites in Models I, II, III, and GS, respectively, with area under the curve (AUC) values ranging between 0.82 and 0.91 in the training cohort and between 0.77 and 0.86 in the validation cohort. The combination of the metabolomic panels and five baseline clinical factors that include serum prostate-specific antigen, age, family history of PC, previously negative biopsy, and abnormal digital rectal examination results significantly increased AUCs (range 0.88–0.91). At 90% sensitivity (validation cohort), 33%, 34%, 41%, and 36% of unnecessary biopsies were avoided in Models I, II, III, and GS, respectively. The above results were successfully validated using LC-MS with the C18 column. Conclusions Urinary metabolomic profiles with baseline clinical factors may accurately predict sPC in men with elevated risk before biopsy.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Purpose: Biomarkers predicting clinically significant prostate cancer (sPC) before biopsy are currently lacking. This study aimed to develop a non-invasive urine test to predict sPC in at-risk men using urinary metabolomic profiles. Materials and Methods: Urine samples from 934 at-risk subjects and 268 treatment-naïve PC patients were subjected to liquid chromatography/mass spectrophotometry (LC-MS)-based metabolomics profiling using both C18 and hydrophilic interaction liquid chromatography (HILIC) column analyses. Four models were constructed (training cohort [n=647]) and validated (validation cohort [n=344]) for different purposes. Model I differentiates PC from benign cases. Models II, III, and a Gleason score model (model GS) predict sPC that is defined as National Comprehensive Cancer Network (NCCN)-categorized favorable-intermediate risk group or higher (Model II), unfavorable-intermediate risk group or higher (Model III), and GS ≥7 PC (model GS), respectively. The metabolomic panels and predicting models were constructed using logistic regression and Akaike information criterion. Results: The best metabolomic panels from the HILIC column include 25, 27, 28 and 26 metabolites in Models I, II, III, and GS, respectively, with area under the curve (AUC) values ranging between 0.82 and 0.91 in the training cohort and between 0.77 and 0.86 in the validation cohort. The combination of the metabolomic panels and five baseline clinical factors that include serum prostate-specific antigen, age, family history of PC, previously negative biopsy, and abnormal digital rectal examination results significantly increased AUCs (range 0.88–0.91). At 90% sensitivity (validation cohort), 33%, 34%, 41%, and 36% of unnecessary biopsies were avoided in Models I, II, III, and GS, respectively. The above results were successfully validated using LC-MS with the C18 column. Conclusions Urinary metabolomic profiles with baseline clinical factors may accurately predict sPC in men with elevated risk before biopsy.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Purpose: Biomarkers predicting clinically significant prostate cancer (sPC) before biopsy are currently lacking. This study aimed to develop a non-invasive urine test to predict sPC in at-risk men using urinary metabolomic profiles. Materials and Methods: Urine samples from 934 at-risk subjects and 268 treatment-naïve PC patients were subjected to liquid chromatography/mass spectrophotometry (LC-MS)-based metabolomics profiling using both C18 and hydrophilic interaction liquid chromatography (HILIC) column analyses. Four models were constructed (training cohort [n=647]) and validated (validation cohort [n=344]) for different purposes. Model I differentiates PC from benign cases. Models II, III, and a Gleason score model (model GS) predict sPC that is defined as National Comprehensive Cancer Network (NCCN)-categorized favorable-intermediate risk group or higher (Model II), unfavorable-intermediate risk group or higher (Model III), and GS ≥7 PC (model GS), respectively. The metabolomic panels and predicting models were constructed using logistic regression and Akaike information criterion. Results: The best metabolomic panels from the HILIC column include 25, 27, 28 and 26 metabolites in Models I, II, III, and GS, respectively, with area under the curve (AUC) values ranging between 0.82 and 0.91 in the training cohort and between 0.77 and 0.86 in the validation cohort. The combination of the metabolomic panels and five baseline clinical factors that include serum prostate-specific antigen, age, family history of PC, previously negative biopsy, and abnormal digital rectal examination results significantly increased AUCs (range 0.88–0.91). At 90% sensitivity (validation cohort), 33%, 34%, 41%, and 36% of unnecessary biopsies were avoided in Models I, II, III, and GS, respectively. The above results were successfully validated using LC-MS with the C18 column. Conclusions Urinary metabolomic profiles with baseline clinical factors may accurately predict sPC in men with elevated risk before biopsy.