Yazar "Aghaei, Mohammadreza" seçeneğine göre listele

Listeleniyor 1 - 6 / 6
  • Küçük Resim Yok
    Öğe
    Accuracy is not enough: explainable boosting machine model and identification of candidate biomarkers for real-time sepsis risk assessment in the emergency department
    (Bmc, 2025) Yagin, Fatma Hilal; Aygun, Umran; Colak, Cemil; Alkhalifa, Amal K.; Alzakari, Sarah A.; Aghaei, Mohammadreza
    BackgroundSepsis poses a significant threat in emergency settings, necessitating tools for early and interpretable risk assessment. This study aimed to develop a robust explainable boosting machine (EBM) model, one of the explainable artificial intelligence (XAI) technologies, to construct a predictive model that balances high accuracy and clinical interpretability for use in emergency departments (EDs) and to examine candidate biomarkers.MethodsThe study identified a significant class imbalance problem in the sepsis distribution among 560 sepsis and 1012 non-sepsis patients. To address the imbalance issue, SMOTE-NC was applied in the training data. The data was divided into two parts, 80% training and 20% testing. To ensure the reliability of the models and to report unbiased results, this process was repeated 100 times and the average performance was reported. To determine the best model for sepsis prediction, five different models (AdaBoost, Gradient Boosting, CatBoost, LightGBM, and EBM) were trained, and their performances were evaluated. In the last stage, we presented local and global explanations of EBM.ResultsThe EBM model achieved the highest success by reaching 79.1% F1-score, 80.9% sensitivity, and 84.8% AUC after resampling. In the global explanations, the variables with the highest weights in the model's decision process were identified as positive blood culture, oxygen saturation, and procalcitonin, respectively.ConclusionThe EBM model accurately predicts sepsis risk based on clinically relevant biomarkers. The model's high performance and inherent transparency can foster trust among clinicians and facilitate its integration into emergency department workflows for real-time decision support.
  • Küçük Resim Yok
    Öğe
    Identification of a Novel Lipidomic Biomarker for Hepatocyte Carcinoma Diagnosis: Advanced Boosting Machine Learning Techniques Integrated with Explainable Artificial Intelligence
    (Mdpi, 2025) Yagin, Fatma Hilal; Colak, Cemil; Al-Hashem, Fahaid; Alzakari, Sarah A.; Alhussan, Amel Ali; Aghaei, Mohammadreza
    Background: Hepatocellular carcinoma (HCC) is a leading cause of cancer-related mortality worldwide, often diagnosed at late stages due to the limited sensitivity of current screening tools. This study explores whether blood-based lipidomic profiling, combined with explainable artificial intelligence (XAI), can improve early and interpretable detection of HCC. Methods: We analyzed lipidomic data from 219 HCC patients and 219 matched healthy controls using liquid chromatography-mass spectrometry. An Explainable Boosting Machine (EBM) was employed to identify discriminatory lipid biomarkers and was compared against several standard machine learning algorithms. Results: The EBM model achieved superior performance with 87.0% accuracy, 87.7% sensitivity, 86.3% specificity, and an AUC of 91.8%, outperforming other models. Key lipid biomarkers identified included specific phosphatidylcholines (PC 38:2, PC 40:4), sphingomyelins (SM d40:2 B), and lysophosphatidylcholines (LPC 18:2), which exhibited significant alterations in HCC patients and highlighted disruptions in sphingolipid metabolism. Conclusions: Integration of lipidomics with explainable machine learning offers a powerful, transparent approach for HCC biomarker discovery, achieving high diagnostic accuracy while providing biological insights. This strategy holds promise for developing non-invasive, clinically interpretable screening tools to improve early detection of liver cancer.
  • Küçük Resim Yok
    Öğe
    Identification of metabolomics-based biomarker discovery in individuals with down syndrome utilizing kernel-tree model-enhanced explainable artificial intelligence methodology
    (Frontiers Media Sa, 2025) Colak, Cemil; Yagin, Fatma Hilal; Yagin, Burak; Alkhateeb, Abedalrhman; Al-Rawi, Mahmood Basil A.; Akhloufi, Moulay A.; Aghaei, Mohammadreza
    Objective: This study aims to develop an explainable artificial intelligence (XAI) model integrated with machine learning (ML) to comprehensively investigate metabolic differences between individuals with Down syndrome (T21) and healthy controls (D21) and to identify novel/pathway-specific biomarkers. In this study, ML classifiers including AdaBoost, LightGBM, Random Forest, KTBoost, and XGBoost are applied to metabolomics data obtained from metabolomic analyses by high-resolution liquid chromatography-mass spectrometry (LC-MS) using blood plasma samples of 316 T21 and 103 D21 individuals, and the importance of metabolites is evaluated by XAI-based SHAP analysis. The KTBoost model shows the highest classification performance with an accuracy of 90.4% and area under the curve (AUC) of 95.9%, outperforming AdaBoost, LightGBM, Random Forest, and XGBoost. Significant downregulation and upregulation of some metabolites were observed in the T21 group compared to the D21 group. Metabolites such as vitamin C, taurolithocholic acid, sphingosine, and prostaglandin A2/B2/J2 are observed at low levels in the T21 group. In contrast, metabolites such as thymidine, tau-roursodeoxycholic acid, serine, and nervonic acid are elevated. SHAP analysis revealed that L-Citrulline, Kynurenin, Prostaglandin A2/B2/J2, Urate, and Pantothenate metabolites could be novel/pathway-specific biomarkers to differentiate the T21 group. This study revealed significant metabolic alterations in individuals with T21 and demonstrated the effectiveness of the combination of ML and XAI methods to identify novel/pathway-specific biomarkers. The findings may contribute to a better understanding of Down syndrome's molecular mechanisms and the development of future diagnostic and therapeutic strategies.
  • Küçük Resim Yok
    Öğe
    Leveraging Explainable Automated Machine Learning (AutoML) and Metabolomics for Robust Diagnosis and Pathophysiological Insights in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS)
    (Mdpi, 2025) Yagin, Fatma Hilal; Colak, Cemil; Al-Hashem, Fahaid; Alzakari, Sarah A.; Alhussan, Amel Ali; Aghaei, Mohammadreza
    Background/Objectives: Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a debilitating complex disease with an elusive etiology, lacking objective diagnostic biomarkers. This study leverages advanced Automated Machine Learning (AutoML) to analyze plasma metabolomic and lipidomic profiles for the purpose of ME/CFS detection. Methods: We utilized a publicly available dataset comprising 888 metabolic features from 106 ME/CFS patients and 91 matched controls. Three AutoML frameworks-TPOT, Auto-Sklearn, and H2O AutoML-were benchmarked under identical time constraints. Univariate ROC and PLS-DA analyses with cross-validation, permutation testing, and VIP-based feature selection were applied to standardized, log-transformed omics data to identify significant discriminatory metabolites/lipids and assess their intercorrelations. Results: TPOT significantly outperformed its counterparts, achieving an area under the curve (AUC) of 92.1%, accuracy of 87.3%, sensitivity of 85.8%, and specificity of 89.0%. The PLS-DA model revealed a moderate but statistically significant discrimination between ME/CFS and controls. Explainable artificial intelligence (XAI) via SHAP analysis of the optimal TPOT model identified key metabolites implicating dysregulated pathways in mitochondrial energy metabolism (succinic acid, pyruvic acid, leucine), chronic inflammation (prostaglandin D2, 11,12-EET), gut-brain axis communication (glycocholic acid), and cell membrane integrity (pc(35:2)a). Conclusions: Our results demonstrate that TPOT-derived models not only provide a highly accurate and robust diagnostic tool but also yield biologically interpretable insights into the pathophysiology of ME/CFS, highlighting its potential for clinical decision support and elucidating novel therapeutic targets.
  • Küçük Resim Yok
    Öğe
    Machine Learning Classification of Cognitive Status in Community-Dwelling Sarcopenic Women: A SHAP-Based Analysis of Physical Activity and Anthropometric Factors
    (Mdpi, 2025) Gormez, Yasin; Yagin, Fatma Hilal; Aygun, Yalin; Alzakari, Sarah A.; Alhussan, Amel Ali; Aghaei, Mohammadreza
    Background and Objectives: Sarcopenia, characterized by progressive loss of skeletal muscle mass and function, has increasingly been recognized not only as a physical health concern but also as a potential risk factor for cognitive decline. This study investigates the application of machine learning algorithms to classify cognitive status based on Mini-Mental State Examination (MMSE) scores in community-dwelling sarcopenic women. Materials and Methods: A dataset of 67 participants was analyzed, with MMSE scores categorized into severe (<= 17) and mild (>17) cognitive impairment. Eight classification models-MLP, CatBoost, LightGBM, XGBoost, Random Forest (RF), Gradient Boosting (GB), Logistic Regression (LR), and AdaBoost-were evaluated using a repeated holdout strategy over 100 iterations. Hyperparameter optimization was performed via Bayesian optimization, and model performance was assessed using metrics including weighted F1-score (w_f1), accuracy, precision, recall, PR-AUC, and ROC-AUC. Results: Among the models, CatBoost achieved the highest w_f1 (87.05 +/- 2.85%) and ROC-AUC (90 +/- 5.65%), while AdaBoost and GB showed superior PR-AUC scores (92.49% and 91.88%, respectively), indicating strong performance in handling class imbalance and threshold sensitivity. SHAP (SHapley Additive exPlanations) analysis revealed that moderate physical activity (moderatePA minutes), walking days, and sitting time were among the most influential features, with higher physical activity associated with reduced risk of cognitive impairment. Anthropometric factors such as age, BMI, and weight also contributed significantly. Conclusions: The results highlight the effectiveness of boosting-based models in capturing complex patterns in clinical data and provide interpretable evidence supporting the role of modifiable lifestyle factors in cognitive health. These findings suggest that machine learning, combined with explainable AI, can enhance risk assessment and inform targeted interventions for cognitive decline in older women.
  • Küçük Resim Yok
    Öğe
    Machine Learning-Integrated Explainable Artificial Intelligence Approach for Predicting Steroid Resistance in Pediatric Nephrotic Syndrome: A Metabolomic Biomarker Discovery Study
    (Mdpi, 2025) Yagin, Fatma Hilal; Inceoglu, Feyza; Colak, Cemil; Alkhalifa, Amal K.; Alzakari, Sarah A.; Aghaei, Mohammadreza
    Aim: Nephrotic syndrome (NS) represents a complex glomerular disorder with significant clinical heterogeneity across pediatric and adult populations. Although glucocorticosteroids have constituted the mainstay of therapeutic intervention for more than six decades, primary treatment resistance manifests in approximately 20% of pediatric patients and 50% of adult cohorts. Steroid-resistant nephrotic syndrome (SRNS) is associated with substantially greater morbidity compared to steroid-sensitive nephrotic syndrome (SSNS), characterized by both iatrogenic glucocorticoid toxicity and progressive nephron loss with attendant decline in renal function. Based on this, the current study aims to develop a robust machine learning (ML) model integrated with explainable artificial intelligence (XAI) to distinguish SRNS and identify important biomarker candidate metabolites. Methods: In the study, biomarker candidate compounds obtained from proton nuclear magnetic resonance (1 H NMR) metabolomics analyses on plasma samples taken from 41 patients with NS (27 SSNS and 14 SRNS) were used. We developed ML models to predict steroid resistance in pediatric NS using metabolomic data. After preprocessing with MICE-LightGBM imputation for missing values (<30%) and standardization, the dataset was randomly split into training (80%) and testing (20%) sets, repeated 100 times for robust evaluation. Four supervised algorithms (XGBoost, LightGBM, AdaBoost, and Random Forest) were trained and evaluated using AUC, sensitivity, specificity, F1-score, accuracy, and Brier score. XAI methods including SHAP (for global feature importance and model interpretability) and LIME (for individual patient-level explanations) were applied to identify key metabolomic biomarkers and ensure clinical transparency of predictions. Results: Among four ML algorithms evaluated, Random Forest demonstrated superior performance with the highest accuracy (0.87 +/- 0.12), sensitivity (0.90 +/- 0.18), AUC (0.92 +/- 0.09), and lowest Brier score (0.20 +/- 0.03), followed by LightGBM, AdaBoost, and XGBoost. The superiority of the Random Forest model was confirmed by paired t-tests, which revealed significantly higher AUC and lower Brier scores compared to all other algorithms (p < 0.05). SHAP analysis identified key metabolomic biomarkers consistently across all models, including glucose, creatine, 1-methylhistidine, homocysteine, and acetone. Low glucose and creatine levels were positively associated with steroid resistance risk, while higher propylene glycol and carnitine concentrations increased SRNS probability. LIME analysis provided patient-specific interpretability, confirming these metabolomic patterns at individual level. The XAI approach successfully identified clinically relevant metabolomic signatures for predicting steroid resistance with high accuracy and interpretability. Conclusions: The present study successfully identified candidate metabolomic biomarkers capable of predicting SRNS prior to treatment initiation and elucidating critical molecular mechanisms underlying steroid resistance regulation.