Amin Mansoori
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Amin Mansoori has a total of 75 co-authors. Here are the top co-authors by number of shared publications:
- Sohrab Effati
Department of Applied Mathematics, Faculty of Mathematical science, Ferdowsi University of Mashhad, Mashhad, Iran
9 shared publications Mohammad Eshaghnezhad
Ferdowsi University of Mashhad
6 shared publications- Habibollah Esmaily
Department of Biostatistics, School of Health, Mashhad University of Medical Sciences, Mashhad, Iran
5 shared publications Majid Ghayour-Mobarhan
Metabolic Syndrome Research Center, Mashhad University of Medical Sciences, Mashhad, Iran
4 shared publicationsMark Ghamsary
School of Public Health, Loma Linda University, Loma Linda, USA
4 shared publicationsGordon Ferns
Brighton and Sussex Medical School, Division of Medical Education, Brighton, UK
3 shared publications
Latest Publications
- 23 September 2026
- Published by Springer Nature in SN Comprehensive Clinical Medicine
- Vol. 8 (1) , 449
Hematological Reference Intervals in Adults from the MASHAD Cohort: A Secondary Analysis Comparing Direct and Indirect Methods
- 23 September 2026
- Published by Springer Nature in SN Comprehensive Clinical Medicine
- Vol. 8 (1) , 449
Hematological Reference Intervals in Adults from the MASHAD Cohort: A Secondary Analysis Comparing Direct and Indirect Methods
- Open Access
- 20 June 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
A machine learning-based framework for predicting hypertension using serum hematological factors
Hypertension (HTN) is a leading global cause of cardiovascular disease (CVD) and all-cause mortality, underscoring the need for early detection and intervention. This study aimed to develop a machine learning (ML)-based predictive framework for HTN using routinely available hematologic and clinical parameters. We analyzed data from the Mashhad Stroke and Heart Atherosclerotic Disorder (MASHAD) study (2010–2020). From an initial 9,704 participants, 4,923 individuals (3,033 without and 1,890 with incident HTN) were included after applying predefined inclusion/exclusion criteria and preprocessing. Predictors included hematologic biomarkers (WBC, RBC, MCV, PLT, RDW, PDW, NLR) and clinical/demographic variables (age, sex, smoking, BMI, GFR, physical activity). Missing values (< 10% for most variables) were addressed via median imputation for continuous features. We employed multiple ML algorithms (XGBoost, LightGBM, logistic regression, decision trees, random forests, gradient boosting, k-nearest neighbors, naive Bayes, ExtraTrees, AdaBoost) to develop and compare models. Following model training and performance-based ranking, XGBoost emerged as the top-performing classifier, achieving a ROC-AUC of 0.66 (95% CI: 0.63–0.69). Feature importance analysis identified age, BMI, RBC count, and MCV as the most influential predictors, with consistent rankings across other models. All models exhibited stable performance across training and test sets, indicating minimal overfitting. Our findings demonstrate an ML-driven framework exploring potential signals in HTN risk using routinely available clinical and hematologic data. While the predictive performance reflects the inherent challenge of forecasting a multifactorial disease from baseline variables, this work establishes a transparent, interpretable exploratory benchmark and identifies key modifiable predictors for future validation and enhancement in more comprehensive datasets.
- 20 June 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessA machine learning-based framework for predicting hypertension using serum hematological factors
Hypertension (HTN) is a leading global cause of cardiovascular disease (CVD) and all-cause mortality, underscoring the need for early detection and intervention. This study aimed to develop a machine learning (ML)-based predictive framework for HTN using routinely available hematologic and clinical parameters. We analyzed data from the Mashhad Stroke and Heart Atherosclerotic Disorder (MASHAD) study (2010–2020). From an initial 9,704 participants, 4,923 individuals (3,033 without and 1,890 with incident HTN) were included after applying predefined inclusion/exclusion criteria and preprocessing. Predictors included hematologic biomarkers (WBC, RBC, MCV, PLT, RDW, PDW, NLR) and clinical/demographic variables (age, sex, smoking, BMI, GFR, physical activity). Missing values (< 10% for most variables) were addressed via median imputation for continuous features. We employed multiple ML algorithms (XGBoost, LightGBM, logistic regression, decision trees, random forests, gradient boosting, k-nearest neighbors, naive Bayes, ExtraTrees, AdaBoost) to develop and compare models. Following model training and performance-based ranking, XGBoost emerged as the top-performing classifier, achieving a ROC-AUC of 0.66 (95% CI: 0.63–0.69). Feature importance analysis identified age, BMI, RBC count, and MCV as the most influential predictors, with consistent rankings across other models. All models exhibited stable performance across training and test sets, indicating minimal overfitting. Our findings demonstrate an ML-driven framework exploring potential signals in HTN risk using routinely available clinical and hematologic data. While the predictive performance reflects the inherent challenge of forecasting a multifactorial disease from baseline variables, this work establishes a transparent, interpretable exploratory benchmark and identifies key modifiable predictors for future validation and enhancement in more comprehensive datasets.
- Open Access
- 10 May 2026
- Published by Wiley in Endocrinology, Diabetes & Metabolism
- Vol. 9 (3) , e70199
A Comparison of Machine Learning Algorithms for Predicting Hypertension Incidence Based on Cohort Study
Objectives: This study aimed to identify key hypertension (HTN) risk factors using machine learning (ML) models to enhance prediction accuracy. Methods: Data from the Mashhad stroke and heart atherosclerotic disorder (MASHAD) cohort, comprising 8237 baseline normotensive subjects, was analysed over a 10‐year follow‐up, during which 2548 developed HTN. Five ML algorithms—K‐nearest neighbours (KNN), logistic regression (LR), XGBoost (XGB), random forest (RF) and neural networks (NN)—were employed to determine the best prediction model and identify the primary factors influencing HTN development. Results: The analysis revealed that the XGBoost model was the most suitable classifier for predicting HTN, outperforming the other algorithms. It achieved the highest AUC‐ROC value (0.79), accuracy (74%), precision of the negative class (86%) and recall of the positive class (74%). Although the precision of the positive class was 55%, and the recall of the negative class was 73%, the XGBoost model demonstrated acceptable performance. Additionally, the ML methods consistently identified age (0.189), copper (0.146), BMI (0.086), triglycerides (0.052), HDL (0.039), glucose (0.039) and uric acid (0.030) as the most influential risk factors, as ranked by SHAP feature importance based on the XGBoost model. Conclusion: The XGBoost model effectively predicted HTN incidence over 10 years, and age, copper, BMI, triglycerides, HDL, glucose and uric acid were the most significant risk factors. These findings highlight the importance of incorporating ML models into the prediction and prevention of hypertension.
- 10 May 2026
- Published by Wiley in Endocrinology, Diabetes & Metabolism
- Vol. 9 (3) , e70199
Open AccessA Comparison of Machine Learning Algorithms for Predicting Hypertension Incidence Based on Cohort Study
Objectives: This study aimed to identify key hypertension (HTN) risk factors using machine learning (ML) models to enhance prediction accuracy. Methods: Data from the Mashhad stroke and heart atherosclerotic disorder (MASHAD) cohort, comprising 8237 baseline normotensive subjects, was analysed over a 10‐year follow‐up, during which 2548 developed HTN. Five ML algorithms—K‐nearest neighbours (KNN), logistic regression (LR), XGBoost (XGB), random forest (RF) and neural networks (NN)—were employed to determine the best prediction model and identify the primary factors influencing HTN development. Results: The analysis revealed that the XGBoost model was the most suitable classifier for predicting HTN, outperforming the other algorithms. It achieved the highest AUC‐ROC value (0.79), accuracy (74%), precision of the negative class (86%) and recall of the positive class (74%). Although the precision of the positive class was 55%, and the recall of the negative class was 73%, the XGBoost model demonstrated acceptable performance. Additionally, the ML methods consistently identified age (0.189), copper (0.146), BMI (0.086), triglycerides (0.052), HDL (0.039), glucose (0.039) and uric acid (0.030) as the most influential risk factors, as ranked by SHAP feature importance based on the XGBoost model. Conclusion: The XGBoost model effectively predicted HTN incidence over 10 years, and age, copper, BMI, triglycerides, HDL, glucose and uric acid were the most significant risk factors. These findings highlight the importance of incorporating ML models into the prediction and prevention of hypertension.
- 15 October 2025
- Published by Springer Nature in Computational and Applied Mathematics
- Vol. 45 (2) , 1-21
A capable gradient-based RNN for fuzzy quadratic optimization problems
- 15 October 2025
- Published by Springer Nature in Computational and Applied Mathematics
- Vol. 45 (2) , 1-21
A capable gradient-based RNN for fuzzy quadratic optimization problems
- Open Access
- 6 October 2025
- Published by Springer Nature in Journal of Health, Population and Nutrition
- Vol. 44 (1) , 1-21
Investigating the relationship between blood factors and HDL-C levels in the bloodstream using machine learning methods
The study investigates the relationship between blood lipid components and metabolic disorders, specifically high-density lipoprotein cholesterol (HDL-C), which is crucial for cardiovascular health. It uses logistic regression (LR), decision tree (DT), random forest (RF), K-nearest neighbors (KNN), XGBoost (XGB), and neural networks (NN) algorithms to explore how blood factors affect HDL-C levels in the bloodstream. The study involved 9704 participants, categorized into normal and low HDL-C levels. Data was analyzed using a data mining approach such as LR, DT, RF, KNN, XGB, and NN to predict HDL-C measurement. Additionally, DT was used to identify the predictive model for HDL-C measurement. This study identified gender-specific hematological predictors of HDL-C levels using multiple ML models. Logistic regression exhibited the highest performance. NHR and LHR were the most influential predictors in males and females, respectively, with SHAP analysis confirming their critical roles alongside LYM, NEUT, and WBC in HDL-C classification. The results show that blood inflammation plays a role in HDL-C homeostasis. The mechanisms of these relationships are not fully understood, but a complex interplay between inflammation and HDL-C levels as well as cardiometabolic health is evident. These findings support the pathophysiological role of inflammatory pathways in cardiometabolic disorders and provide insights into how modulation of hematological inflammation may contribute to disease prevention or treatment.
- 6 October 2025
- Published by Springer Nature in Journal of Health, Population and Nutrition
- Vol. 44 (1) , 1-21
Open AccessInvestigating the relationship between blood factors and HDL-C levels in the bloodstream using machine learning methods
The study investigates the relationship between blood lipid components and metabolic disorders, specifically high-density lipoprotein cholesterol (HDL-C), which is crucial for cardiovascular health. It uses logistic regression (LR), decision tree (DT), random forest (RF), K-nearest neighbors (KNN), XGBoost (XGB), and neural networks (NN) algorithms to explore how blood factors affect HDL-C levels in the bloodstream. The study involved 9704 participants, categorized into normal and low HDL-C levels. Data was analyzed using a data mining approach such as LR, DT, RF, KNN, XGB, and NN to predict HDL-C measurement. Additionally, DT was used to identify the predictive model for HDL-C measurement. This study identified gender-specific hematological predictors of HDL-C levels using multiple ML models. Logistic regression exhibited the highest performance. NHR and LHR were the most influential predictors in males and females, respectively, with SHAP analysis confirming their critical roles alongside LYM, NEUT, and WBC in HDL-C classification. The results show that blood inflammation plays a role in HDL-C homeostasis. The mechanisms of these relationships are not fully understood, but a complex interplay between inflammation and HDL-C levels as well as cardiometabolic health is evident. These findings support the pathophysiological role of inflammatory pathways in cardiometabolic disorders and provide insights into how modulation of hematological inflammation may contribute to disease prevention or treatment.
- Open Access
- 22 January 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1
Publisher Correction: Predicting high sensitivity C-reactive protein levels and their associations in a large population using decision tree and linear regression
- 22 January 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1
Open AccessPublisher Correction: Predicting high sensitivity C-reactive protein levels and their associations in a large population using decision tree and linear regression
- Open Access
- 5 December 2024
- Published by Springer Nature in Scientific Reports
- Vol. 14 (1) , 1-16
Predicting high sensitivity C-reactive protein levels and their associations in a large population using decision tree and linear regression
High-sensitivity C-reactive protein (hs-CRP) is a biomarker of inflammation predicting the incidence of different health pathologies. In this study, we aimed to evaluate the association between hematological and demographic factors with hs-CRP levels using decision tree (DT) and linear regression (LR) modeling. This study was conducted on a population of 9704 males and females aged 35 to 65 years recruited from the Mashhad Stroke and Heart Atherosclerotic Disorder (MASHAD) cohort study. We utilized a data mining approach to construct a predictive model of hs-CRP measurements, employing the DT methodology. DT model was used to predict hs-CRP level using biochemical factors and clinical features. A total of 9,704 individuals were included in the analysis, with 57% of them being female. Except for fasting blood glucose (FBG), hypertension (HTN), and Type 2 diabetes mellites (T2DM), all variables showed significant differences between the two groups. The results of the LR models showed that variables such as anxiety score, depression score, Systolic Blood Pressure, Cardiovascular disease, and HTN were significant in predicting hs-CRP levels. In the DT models, depression score, FBG, cholesterol, and anxiety score were identified as the most important factors in predicting hs-CRP levels. DT model was able to predict hs-CRP level with an accuracy of 72.1% in training and 71.4% in testing of both genders. The proposed DT model appears to be able to predict the hs-CRP levels based on anxiety score, depression scores, fasting blood glucose, systolic blood pressure, and history of cardiovascular diseases.
- 5 December 2024
- Published by Springer Nature in Scientific Reports
- Vol. 14 (1) , 1-16
Open AccessPredicting high sensitivity C-reactive protein levels and their associations in a large population using decision tree and linear regression
High-sensitivity C-reactive protein (hs-CRP) is a biomarker of inflammation predicting the incidence of different health pathologies. In this study, we aimed to evaluate the association between hematological and demographic factors with hs-CRP levels using decision tree (DT) and linear regression (LR) modeling. This study was conducted on a population of 9704 males and females aged 35 to 65 years recruited from the Mashhad Stroke and Heart Atherosclerotic Disorder (MASHAD) cohort study. We utilized a data mining approach to construct a predictive model of hs-CRP measurements, employing the DT methodology. DT model was used to predict hs-CRP level using biochemical factors and clinical features. A total of 9,704 individuals were included in the analysis, with 57% of them being female. Except for fasting blood glucose (FBG), hypertension (HTN), and Type 2 diabetes mellites (T2DM), all variables showed significant differences between the two groups. The results of the LR models showed that variables such as anxiety score, depression score, Systolic Blood Pressure, Cardiovascular disease, and HTN were significant in predicting hs-CRP levels. In the DT models, depression score, FBG, cholesterol, and anxiety score were identified as the most important factors in predicting hs-CRP levels. DT model was able to predict hs-CRP level with an accuracy of 72.1% in training and 71.4% in testing of both genders. The proposed DT model appears to be able to predict the hs-CRP levels based on anxiety score, depression scores, fasting blood glucose, systolic blood pressure, and history of cardiovascular diseases.
- Open Access
- 29 April 2024
- Published by Springer Nature in BMC Endocrine Disorders
- Vol. 24 (1) , 1-8
The associations of oxidative stress and inflammatory markers with obesity in Iranian population: MASHAD cohort study
Low-grade inflammation and stress oxidative condition play a role in the pathogenesis of obesity, and the serum levels of these markers, such as pro-oxidant-antioxidant balance (PAB), high-sensitivity C-reactive protein (hs-CRP), and uric acid may indicate obesity progression. In this study, we aimed to investigate the relationship between obesity with PAB, hs-CRP, and uric acid in the Iranian population. This study was derived from the Mashhad Stroke and Heart Atherosclerotic Disorder (MASHAD) study. A total of 7985 subjects aged 35 to 65 years were divided into three groups according to body mass index (BMI) as: normal, overweight and obese groups. Anthropometric indices and biochemical parameters such as PAB, superoxide dismutase type 1 (SOD1), hs-CRP, and uric acid were measured in all the participants. We evaluated the association of obesity with inflammatory factors by using multivariate regression analysis. Also, those participants with hypertension, an endocrine disorder, history of cardiovascular diseases and diabetes mellitus were excluded from the study. There was a positive significant correlation between BMI and serum PAB, hs-CRP and uric acid (p < 0.001). While no statistically significant relation was observed between BMI and SOD1 (p = 0.85). Multivariate regression analysis showed that the risk of overweight and obesity increased 1.02 and 1.03-fold according to increase 10 units of PAB raise in comparison to reference group (normal weight) [(odds ratio (OR): 1.02, 95% CI (1.01–1.03)] and [OR: 1.03, 95% CI (1.01–1.04)], respectively). In addition, hs-CRP serum concentration was significantly associated with a high risk of obesity [(OR: 1.02; 95% CI (1.01–1.03)]. While the high levels of serum uric acid were associated with increased odds of overweight and obesity risk [OR: 1.4; CI (1.39–1.58) and OR: 1.76; CI (1.63–1.89), respectively]. Generally, we showed a significant association between BMI and serum PAB, hs-CRP values and uric acid levels, suggesting the role of these factors as risk stratification factors for obesity.
- 29 April 2024
- Published by Springer Nature in BMC Endocrine Disorders
- Vol. 24 (1) , 1-8
Open AccessThe associations of oxidative stress and inflammatory markers with obesity in Iranian population: MASHAD cohort study
Low-grade inflammation and stress oxidative condition play a role in the pathogenesis of obesity, and the serum levels of these markers, such as pro-oxidant-antioxidant balance (PAB), high-sensitivity C-reactive protein (hs-CRP), and uric acid may indicate obesity progression. In this study, we aimed to investigate the relationship between obesity with PAB, hs-CRP, and uric acid in the Iranian population. This study was derived from the Mashhad Stroke and Heart Atherosclerotic Disorder (MASHAD) study. A total of 7985 subjects aged 35 to 65 years were divided into three groups according to body mass index (BMI) as: normal, overweight and obese groups. Anthropometric indices and biochemical parameters such as PAB, superoxide dismutase type 1 (SOD1), hs-CRP, and uric acid were measured in all the participants. We evaluated the association of obesity with inflammatory factors by using multivariate regression analysis. Also, those participants with hypertension, an endocrine disorder, history of cardiovascular diseases and diabetes mellitus were excluded from the study. There was a positive significant correlation between BMI and serum PAB, hs-CRP and uric acid (p < 0.001). While no statistically significant relation was observed between BMI and SOD1 (p = 0.85). Multivariate regression analysis showed that the risk of overweight and obesity increased 1.02 and 1.03-fold according to increase 10 units of PAB raise in comparison to reference group (normal weight) [(odds ratio (OR): 1.02, 95% CI (1.01–1.03)] and [OR: 1.03, 95% CI (1.01–1.04)], respectively). In addition, hs-CRP serum concentration was significantly associated with a high risk of obesity [(OR: 1.02; 95% CI (1.01–1.03)]. While the high levels of serum uric acid were associated with increased odds of overweight and obesity risk [OR: 1.4; CI (1.39–1.58) and OR: 1.76; CI (1.63–1.89), respectively]. Generally, we showed a significant association between BMI and serum PAB, hs-CRP values and uric acid levels, suggesting the role of these factors as risk stratification factors for obesity.
- 27 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control of Network Systems
- Vol. 11 (3) , 1262-1274
Continuous-Time Distributed Generalized Nash Equilibrium Seeking in Nonsmooth Fuzzy Aggregative Games
This paper explores the variational generalized Nash equilibrium seeking for a class of fuzzy aggregative games, where each player's cost function is fuzzy and nonsmooth, and the strategy profile is constrained by a coupling nonlinear inequality, a affine coupling equality with $\alpha$ -cuts and heterogeneous local convex sets. Under the differential inclusion framework, a continuous-time distributed algorithm with derivative feedback term is proposed. Benefited from this algorithm, each player is assigned several auxiliary variables to share with its neighbors and estimate the aggregate function, so that the important information such as cost function, constrains and decision are privatized. Based on fuzzy optimization and nonsmooth analysis, the asymptotic convergence of the algorithm to generalized Nash equilibrium is rigorously proved, and the convergence is not affected by the initial strategy profile of the players. Moreover, two application instances are used to demonstrate the theoretical results.
- 27 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control of Network Systems
- Vol. 11 (3) , 1262-1274
Continuous-Time Distributed Generalized Nash Equilibrium Seeking in Nonsmooth Fuzzy Aggregative Games
This paper explores the variational generalized Nash equilibrium seeking for a class of fuzzy aggregative games, where each player's cost function is fuzzy and nonsmooth, and the strategy profile is constrained by a coupling nonlinear inequality, a affine coupling equality with $\alpha$ -cuts and heterogeneous local convex sets. Under the differential inclusion framework, a continuous-time distributed algorithm with derivative feedback term is proposed. Benefited from this algorithm, each player is assigned several auxiliary variables to share with its neighbors and estimate the aggregate function, so that the important information such as cost function, constrains and decision are privatized. Based on fuzzy optimization and nonsmooth analysis, the asymptotic convergence of the algorithm to generalized Nash equilibrium is rigorously proved, and the convergence is not affected by the initial strategy profile of the players. Moreover, two application instances are used to demonstrate the theoretical results.