Yazar "Yilmaz, Murat" seçeneğine göre listele

Listeleniyor 1 - 3 / 3
  • Küçük Resim Yok
    Öğe
    Forecasting Stone-Free Status Following Percutaneous Nephrolithotomy Utilizing Explainable Machine Learning
    (Mdpi, 2026) Cicek, Resul; Topcu, Ibrahim; Dural, Bulut; Balikci Cicek, Ipek; Yilmaz, Murat; Colak, Cemil
    Background: This study aimed to create and evaluate explainable machine learning models for forecasting postoperative stone-free status following percutaneous nephrolithotomy (PNL) utilizing a substantial clinical cohort. Methods: This retrospective single-center analysis encompassed 2144 adult patients who received PNL from 2010 to 2024. We employed clinical, radiographic, stone-related, and surgical data to train four supervised machine learning models: Extreme Gradient Boosting (XGBoost), Random Forest, Light Gradient Boosting Machine (LightGBM), and Adaptive Boosting (AdaBoost). We used the Synthetic Minority Oversampling Technique exclusively on the training set to fix the class imbalance. We assessed the model's accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (ROC-AUC) to see how well it worked. SHapley Additive exPlanations (SHAP) were used to measure explainability. Results: The total stone-free rate was 84.8%. XGBoost had the best predictive performance of the models tested, with an accuracy of 0.916 and a ROC-AUC of 0.975. LightGBM was close behind. Random Forest and AdaBoost had relatively inferior performance. SHAP analysis identified anatomical anomalies as demonstrated the strongest association with stone-free outcomes. The size of the access sheath and the number of stones were next. Other parameters that were identified by SHAP as important contributors to model predictions were the placement of the stone, Guy's Stone Score, the length of the operation, and the density of the stone. These feature associations demonstrated clinical coherence with established knowledge in surgical practice. Conclusions: Explainable machine learning algorithms, especially XGBoost, can accurately predict stone-free outcomes following PNL in a way that makes sense to doctors. The incorporation of SHAP improves transparency and facilitates the prospective application of these models as decision-support instruments in personalized surgical planning.
  • Küçük Resim Yok
    Öğe
    Impact of telephonic interviews on persistence and daily adherence to insulin treatment in insulin-naive type 2 diabetes patients: dropout study
    (Dove Medical Press Ltd, 2016) Yavuz, Dilek Gogas; Bilen, Habip; Sancak, Seda; Garip, Tayfun; Hekimsoy, Zeliha; Sahin, Ibrahim; Yilmaz, Murat
    Objective: The objective of this study is to evaluate the impact of sequential telephonic interviews on treatment persistence and daily adherence to insulin injections among insulin-naive type 2 diabetes patients initiated on different insulin regimens in a 3-month period. Methods: A total of 1,456 insulin-naive patients with type 2 diabetes (mean [standard deviation, SD] age: 56.0 [12.0] years, 49.1% were females) initiated on insulin therapy and consecutively randomized to sequential (n=733) and single (n=723) telephonic interview groups were included. Data on insulin treatment and self-reported blood glucose values were obtained via telephone interview. Logistic regression analysis was performed for factors predicting increased likelihood of persistence and skipping an injection. Results: Overall, 76.8% patients (83.2% in sequential vs 70.3% in single interview group, (P<0.001) remained on insulin treatment at the third month. Significantly higher rate for skipping doses was noted in basal bolus than in other regimens (27.0% vs 15.0% for premixed and 15.8% basal insulin, respectively, P<0.0001). Logistic regression analysis revealed sequential telephonic interview (odds ratio [OR], 1.531; 95% confidence interval [CI], 1.093-2.143; P=0.013), higher hemoglobin A1c levels (OR, 1.090; 95% CI, 0.999-1.189; P=0.049), and less negative appraisal of insulin therapy as significant predictors of higher persistence. Basal bolus regimen (OR, 1.583; 95% CI, 1.011-2.479; P=0.045) and higher hemoglobin A1c levels (OR, 1.114; 95% CI, 1.028-1.207; P=0.008) were the significant predictors of increased likelihood of skipping an injection. Conclusion: Our findings revealed positive influence of sequential telephonic interview, although including no intervention in treatment, on achieving better treatment persistence in type 2 diabetes patients initiating insulin.
  • Küçük Resim Yok
    Öğe
    A NEW APPROACH FOR THE PREDICTION OF BRITTLENESS INDEX BASED ON CHEMICAL PROPERTIES OF BASALTIC ROCKS
    (Acad Sci Czech Republic Inst Rock Structure & Mechanics, 2021) Bilen, Candan; Er, Selman; Tugrul, Atiye; Yilmaz, Murat
    Rock brittleness is one of the most important issues in rock mechanics. There is not yet an available method for defining or measuring directly the rock brittleness. The aim of this study is to suggest a new chemical index parameter for the prediction of basaltic rocks' brittleness. In the order of that abovementioned purpose, a total of 23 basaltic rock samples were collected from different region of Turkey. Samples were initially tested to determine their chemical properties. Then, mechanical tests were carried out to define the brittleness indices (B1, B2, and B3) for each corresponding sample. Finally, relations between parameters obtained from test results and brittleness indices were examined with regression analysis. According to the results obtained, a new chemical parameter (CP) was proposed for predicting brittleness via major oxide element components of basaltic rocks. It was found out that, B1 and B2 are not reliable parameters for predicting the different properties, however; B3 and CP can be employed as good criteria for predicting the different properties of basaltic rocks (especially in terms of chemical and mechanical properties).