Mehmet POLAT

Dr. Mehmet POLAT

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 May 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
MÖMurat ÖZDEDE
Murat ÖZDEDE
SYSerap Yadigar
Serap Yadigar
AGAlper Tuna Guven
Alper Tuna Guven
SASuat Akgür
Suat Akgür
FAFelemez ARSLAN
Felemez ARSLAN
EBElif Arı Bakır
Elif Arı Bakır
Background: Finerenone is associated with a lower, yet clinically relevant, risk of hyperkalemia compared with steroidal mineralocorticoid receptor antagonists in diabetic kidney disease (DKD) trials. However, real-world data on hyperkalemia and its associated factors are lacking. Methodology: FINE-TURK is a national, observational cohort of DKD patients who were initiated on finerenone. Eligible adults were included; demographic, clinical, and laboratory data were evaluated. The primary outcome (PO) was hyperkalemia risk signal (potassium ≥ 5.0 mEq/L), and the secondary outcome (SO) was clinically meaningful hyperkalemia (potassium ≥ 5.5 mEq/L). Multivariate logistic regression (LR) was used to define features associated with both PO and SO. LR, random forest (RF), gradient boosting, and CatBoost classifiers were used to define important features associated with the PO. Results: 699 patients were included. 259 (37.1%) reached the PO, and 51 (7.3%) reached the SO. Baseline potassium and estimated glomerular filtration rate (eGFR) were the most important variables associated with both outcomes and were consistently identified as the top features across all models. Thiazide use, presence of diabetic retinopathy, and diabetes duration were also associated with the PO. LR demonstrated the highest recall; random forest achieved the highest precision in performance. Discussion: Real-world data suggest that the risk of clinically meaningful hyperkalemia is similar to that in the clinical trials. In parallel with the safety analysis of clinical trials, baseline potassium and eGFR were consistently the most important factors associated with hyperkalemia risk.

30 June 2026 in European Journal of Theoretical and Applied Sciences
Verified
Journal Article
Open access
PDF file
MDMustafa Demirci
Mustafa Demirci
ÖSÖznur Sari
Öznur Sari
MPMehmet POLAT
Mehmet POLAT
PKPinar Koçatakan
Pinar Koçatakan
Unlike classical Starling hydrostatic models, this study treats the glomerular filtration barrier (GFB) as an open thermodynamic system operating far from equilibrium. Aiming to overcome the delayed macro- outputs of traditional clearance biomarkers (creatinine and urea), this research introduces the "Glomerular Entropy Score" (Sglom) index to literature in order to instantaneously model real-time structural and ionic/osmotic disorder at the membrane level. Retrospectively analyzed records of 544 patients (N=544) were analyzed based on an empirical biophysical model across three independent software pipelines (Orange, RapidMiner, and JASP): Orange (3.38): The predictive power for long-term permanent renal damage was evaluated using Multilayer Perceptron Neural Network (MLP-NN), Naïve Bayes, Logistic Regression, and Random Forest (RF) algorithms; RapidMiner Studio (10.2): Model generalizability was verified using an isolated feature subset and 10-fold cross-validation; JASP (0.98.1): Descriptive and inferential biostatistical analyses, Student's t-test, and hypothesis validation via parametric matrices were conducted (JASP Team, 2024). Findings demonstrated that the calculated entropy index exhibits a non-linear, rightward-expanding chaotic tail structure against peak creatinine values. This condition proves thermodynamic disorder beyond critical bifurcation thresholds. Box plot analyses confirmed that local entropy production and clinical variation were significantly higher in the group with permanent renal failure (t = 2.149, df = 542, p = 0.032). In the Orange architecture, non-linear RF demonstrated superior performance (AUC = 0.800, F1 = 0.858, MCC = 0.518), while the RapidMiner RF model achieved an overall accuracy of 84.03% and a class sensitivity of 98.65% in predicting cases preserving filtration stability. The integration of non- equilibrium thermodynamic metrics with artificial intelligence and biostatistics offers a novel diagnostic framework for capturing GFB damage at an early stage before functional impairment settles.

29 August 2026 in European Journal of Theoretical and Applied Sciences
Verified
Journal Article
Open access
PDF file
MDMustafa Demirci
Mustafa Demirci
ÖSÖznur Sarı
Öznur Sarı
MPMehmet POLAT
Mehmet POLAT
PKPınar Koçatakan
Pınar Koçatakan
Background: Persistent renal dysfunction after acute kidney injury (AKI) is clinically important; however, recovery trajectories are often summarized using conventional threshold-based measures. We evaluated whether nonlinear machine-learning methods could characterize renal dysfunction after AKI and investigated a quantum tunneling-inspired mathematical analogy not as evidence of a quantum- biological mechanism, but as a conceptual framework. Methods: This retrospective cohort included 544 adults with AKI. Admission and peak biochemical and hematological variables were analyzed using Orange Data Mining (3.38.1), RapidMiner (10.2), and Weka (4.9.4). Derived indices were constructed from routinely obtainable clinical variables; t-SNE visualization, Random Forest, support vector machine (SVM), artificial neural network, logistic regression, and decision-tree analyses were performed. Ten-fold cross-validation was used for the reported machine- learning evaluations. Results: The cohort included 444 patients (81.62%) with recovery and 100 patients (18.38%) with persistent renal dysfunction. In the Orange analysis, Random Forest achieved an AUC of 0.812, compared with 0.649 for SVM and 0.774 for the artificial neural network. In Weka, Random Forest achieved 89.33% accuracy and an AUC of 0.850, whereas logistic regression achieved 75.55% accuracy and an AUC of 0.812. Feature ranking identified admission CRP, WBC, ferritin, and D-dimer among the highest-ranked variables. Accordingly, the quantum tunneling analogy was used solely to describe abrupt or threshold-like nonlinear transitions in the clinical feature space. Conclusion: Nonlinear machine-learning models identified clinically meaningful structure associated with persistent renal dysfunction after AKI. The proposed quantum tunneling-inspired framework is a mathematical analogy and requires independent validation before any mechanistic interpretation.

15 September 2026 in PROBLEMS OF ENDOCRINE PATHOLOGY
Verified
Journal Article
Open access
PDF file
ÖSÖznur Sari
Öznur Sari
MPMehmet POLAT
Mehmet POLAT
Background. Diabetes mellitus is a chronic, multisystem disease, which needs appropriate clinical records and coding to facilitate safe patient management, quality of measurement, payment, and population health observation. The international change in classification of diseases, Tenth Revision (ICD-10) to the Eleventh Revision (ICD-11) is a significant structural and conceptual move in the classification of diseases, specifically diabetes because of its complicated complication archetypes and longitudinal care pathways. The ICD-10 influenced documentation practices not be sufficient to promote the greater granularity and digit-first architecture of ICD-11. Objective. This paper will seek to put together the available secondary literature on changes in ICD-10 to ICD-11 in diabetes mellitus with respect to its clinical documentation, coding of complications, endocrine workflows, and health informatics systems implications. Materials and methods. It used narrative secondary research design, which entailed thematic synthesis. The types of literature used include peer-reviewed, international guidance documents and publications of health informatics studies without older than 2018 and 2025, selected by using PubMed, Scopus and IEEE Xplore. Sources that are related to ICD-11 implementation, coding of diabetes, clinical documentation, electronic health records, and data governance were incorporated. The thematic synthetization of findings was aimed at determining repeated implications on practice and informatics. Results. Five connected themes were found, which include changing documentation needs in diabetes care, endangering the accuracy of complication coding and case-mix integrity, requiring reformed electronic health record templates and clinical decision support, facing difficulties in data governance, reporting, and quality measurement, and workforce preparedness to adopt ICD-11. In all themes, poor documentation specificity was identified as the key risk to effective implementation of ICD-11. Conclusions. ICD-11 transition has considerable implications in diabetes care beyond the level of the replacement of technical codes. Failure to change documentary practice in a systematic fashion, informatics infrastructure, and training of workforce will likely undermine data quality and clinical insight by ICD-11 adoption. Active synchronization of endocrine record, EHRS, and governance structure is needed in order to achieve the clinical and analytical advantages of ICD-11.

24 June 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
AGAlper Tuna Guven
Alper Tuna Guven
SYSerap Yadigar
Serap Yadigar
MÖMurat ÖZDEDE
Murat ÖZDEDE
SASuat Akgür
Suat Akgür
FAFelemez ARSLAN
Felemez ARSLAN
EBElif Arı Bakır
Elif Arı Bakır
Background/Objectives: Finerenone improves cardiovascular and renal outcomes in diabetic kidney disease (DKD), but hyperkalemia remains a key safety concern. Patients with elevated baseline potassium levels (≥ 4.9 mEq/L) are largely excluded from clinical trials, and real-world data in this population are scarce. Methods: In this retrospective multicenter cohort study derived from the FINE-TURK cohort, adults with DKD who initiated finerenone with baseline serum potassium ≥ 4.9 mEq/L were included. The primary outcome was clinically significant hyperkalemia (CSH) (≥ 5.5 mEq/L) within three months. Multivariable logistic regression analyses were used to identify associated factors, with multiple sensitivity analyses performed. Results: 166 patients were included, of whom 47 (28.3%) had baseline potassium levels between 5.1 to 5.5 mEq/L. Of the 166 patients, 35 (21.1%) developed CSH, and 10 (6%) patients had follow-up potassium ≥ 6.0 mEq/L. 126 (76.8%) patients required no intervention, 24 (14.6%) were initiated on potassium binders, and finerenone was discontinued only in 12 (7.3%) patients. Baseline eGFR, baseline urinary albumin, loop diuretic use and 20mg finerenone dose were associated with CSH. In contrast, baseline serum potassium was not associated with CSH. Conclusions: In patients with DKD and elevated baseline potassium levels, finerenone initiation was associated with manageable rates of hyperkalemia. Our findings support the cautious use of finerenone in selected patients under close monitoring, as well as highlight the need for a multidimensional approach to hyperkalemia risk assessment.

15 September 2025 in PROBLEMS OF ENDOCRINE PATHOLOGY
Verified
Journal Article
Open access
PDF file
MPMehmet POLAT
Mehmet POLAT
SBSeda Basmaz
Seda Basmaz
Background: Adrenal insufficiency causes nonspecific symptoms such as general fatigue, signs such as hypotension, and abnormalities such as hypoglycemia and leads to a poor prognosis. However, all these are also observed in hemodialysis (HD) patients without adrenal insufficiency. This study aims to investigate the prevalence and clinical characteristics of adrenal insufficiency in patients undergoing hemodialysis due to end-stage kidney failure. Aims: This study was performed to clarify the characteristics of adrenal insufficiency in hemodialysis patients. Methods. Between January-June 2020, 243 patients aged ≥18 years who underwent hemodialysis treatment in Karabuk Training and Research Hospital were evaluated. Among these patients, 137 patients who were sent to the endocrine outpatient clinic with a pre-diagnosis of possible adrenal insufficiency were included in the study. Based on baseline cortisol levels and results of the low-dose 1 μg ACTH stimulation test, patients were categorized as having or not having adrenal insufficiency. Parametric data among the groups were compared with Student T-test and non-parametric data were compared with Mann-Whitney U-test. Chi-square test was used to compare categorical data between both groups. Multivariate logistic regression analysis was also used. Results. Eight HD patients were diagnosed with adrenal insufficiency. There was no significant difference between groups in age, gender, BMI, dialysis duration, or Ca×P. However, patients with adrenal insufficiency had significantly lower Kt/V values (median 1.35 vs. 1.5, p = 0.018). A Kt/V value below 1.35 can be considered an independent predictor of adrenal insufficiency (OR: 59.5, 95% CI: 1.402–2524, p = 0.033). Conclusions. Adrenal insufficiency should be considered in hemodialysis patients presenting with non-specific symptoms such as nausea, vomiting, hypotension, and fatigue. A Kt/V value below 1.35 may serve as an independent predictor of adrenal insufficiency in this population.