Mevlut Ture

Comparing performances of logistic regression, classification and regression tree, and neural networks for predicting coronary artery disease

In this study, performances of classification techniques were compared in order to predict the presence of coronary artery disease (CAD). A retrospective analysis was performed in 1245 subjects (865 presence of CAD and 380 absence of CAD). We compared performances of logistic regression (LR), classification and regression tree (CART), multi-layer perceptron (MLP), radial basis function (RBF), and self-organizing feature maps (SOFM). Predictor variables were age, sex, family history of CAD, smoking status, diabetes mellitus, systemic hypertension, hypercholesterolemia, and body mass index (BMI). Performances of classification techniques were compared using ROC curve, Hierarchical Cluster Analysis (HCA), and Multidimensional Scaling (MDS). Areas under the ROC curves are 0.783, 0.753, 0.745, 0.721, and 0.675, respectively for MLP, LR, CART, RBF, and SOFM. MLP was found the best technique to predict presence of CAD in this data set, given its good classificatory performance. MLP, CART, LR, and RBF performed better than SOFM in predicting CAD in according to HCA and MDS.

Comparing classification techniques for predicting essential hypertension

Hypertension is a leading cause of heart disease and stroke. In this study, performance of classification techniques is compared in order to predict the risk of essential hypertension disease. A retrospective analysis was performed in 694 subjects (452 patients and 242 controls). We compared performances of three decision trees, four statistical algorithms, and two neural networks. Predictor variables were age, sex, family history of hypertension, smoking habits, lipoprotein (a), triglyceride, uric acid, total cholesterol, and body mass index (BMI). Classification techniques were grouped using hierarchical cluster analysis (HCA). The data points appeared to cluster in three groups. The first cluster included MLP and RBF. Furthermore CART which was more similar than other techniques linked this cluster. The second cluster included FDA/MARS (degree=1), LR and QUEST, but FDA/MARS (degree=1) and LR was more similar than QUEST. The third cluster included FDA/MARS (degree=2), CHAID and FDA, but FDA/MARS (degree=2) and CHAID was more similar than FDA. MLP and RBF which are one each of neural networks procedures, performed better than other techniques in predicting hypertension. QUEST had a lesser performance than other techniques.

Using Kaplan-Meier analysis together with decision tree methods (C&RT, CHAID, QUEST, C4.5 and ID3) in determining recurrence-free survival of breast cancer patients

Current evidence supports a clear association between clinical and pathologic factors and recurrence-free survival (RFS) in breast cancer patients. The Cox regression model is the most common tool for investigating simultaneously the influence of several factors on the survival time of patients. But it gives no estimate of the degree of separation of the different subgroups. We propose to analyze different decision tree methods (C&RT, CHAID, QUEST, C4.5 and ID3) and use them additionally to the well-known Kaplan-Meier estimates to investigate the predictive power of these methods. Five hundred patients were included to the study. Two hundred and seventy-nine of them had complete data for prognostic factors and median follow-up is about 40.5 months. First, decision tree methods were analyzed for prognostic factors. Then, according to multidimensional scaling method C4.5 (error rate 0.2258 for training set and 0.3259 for cross-validation) performed slightly better than other methods in predicting risk factors for recurrence. Tumor size, age of menarche, hormonal therapy, histological grade and axillary nodal status are found that an important risk factors for the recurrence. Eight terminal nodes were found and stratified by Kaplan-Meier survival curves. Larger tumor size (>=4.4cm) and receiving no hormonal therapy in a small subgroup of patients were associated with worse prognosis. The five-year RFS is 71.3% in the whole patient population. The sensitivity, specificity and predictive rates calculated by C4.5 method were found 43.8%, 91% and 77.4%, respectively. In this study, C4.5 showed a better degree of separation. As a result, we recommend to use decision tree methods together with Kaplan-Meier analysis to determine risk factors and effect of this factors on survival.

Monitoring Indoor Airborne Fungi and Bacteria in the Different Areas of Trakya University Hospital, Edirne, Turkey

The aim of this investigation was to monitor monthly the densities and distribution of indoor airborne fungi and bacteria in 6 different areas of Trakya University Hospital (Edirne, Turkey). Areas monitored were an operating theatre, birthing-room, emergency department, service area for infectious diseases, intensive care unit and the canteen. Our method was to expose Petri dishes which contained rose-bengal streptomycin agar and 5% sheep-blood agar media to room air for 10-min periods. Samples were collected at 1-month intervals from September 2000 to February 2001. A total of 156 microfungal and 535 bacterial colonies were counted on 144 plates. During a 6-month period, 10 bacterial genera (Acinetobacter, Bacillus, Corynebacterium, Enterococcus, Escherichia, Listeria, Micrococcus, Propionibacteria, Staphylococcus and Streptococcus) 7 fungal genera (Alternaria, Aspergillus, Cladosporium, Paecilomyces, Penicillium, Scopulariopsis and Trichothecium) and 33 fungal species were isolated from the hospital air. Penicillium loliense, P. melinii and P.phoeniceum were newly recognised species for Turkey. Some bacterial species such as coagulase-negative Staphylococcus, Micrococcus and Corynebacterium spp. were predominant (percentages of colonies counted were 72.2, 10.7 and 8.8%, respectively). Cladosporium and Penicillium were the most prevalent fungal genera. Cladosporium was predominant in September, November and February, Alternaria in October and December and Penicillium in January. Staphylococcus spp. was the most common bacterial species in all months. Statistical analyses (regression with optimal scaling test) were applied to the data.

Comparison of four different time series methods to forecast hepatitis A virus infection

Abstract Hepatitis A virus (HAV) infection is not a problem of only developing countries, but also of developed countries. In this study, we compared time series prediction capabilities of three artificial neural networks (ANN) algorithms (multi-layer perceptron (MLP), radial basis function (RBF), and time delay neural networks (TDNN)), and auto-regressive integrated moving average (ARIMA) model to HAV forecasting. To assess the effectiveness of these methods, we used in forecasting 13 years of time series (January 1992–June 2004) monthly records for HAV data, in Turkey. Results show that MLP is more accurate and performs better than RBF, TDNN and ARIMA model.

Monitoring of Fungi and Bacteria in the Indoor Air of Primary Schools in Edirne City, Turkey

We monitored levels of bacteria and fungi in the indoor air at selected sites of several public primary schools in the city of Edirne, Turkey. Sampling was by the Petri plate method onto both a Rose-Bengal streptomycin agar medium and a 5% sheep-blood agar medium exposed to the air for 10-minute periods. Samples were collected monthly over a period of 6 months between August 2001 and January 2002. A total of 941 microfungi and 2066 bacterial colonies were counted on 90 Petri plates. During this 6-month period, 19 bacterial genera, 15 fungal genera and 48 species of fungi were isolated from the air in the schools. Some bacteria, such as coagulase-negative Staphylococcus, Corynebacterium and Bacillus, were predominant (42.7%, 20.4% and 6.9% of the total, respectively). Penicillium, Cladosporium and Alternaria were the most common fungal genera (42.8%, 19.3% and 10.1% of the total, respectively). Staphylococcus, Acinetobacter, Corynebacterium, Propionibacterium and Pseudomonas genera were found in every month. Statistical analysis of the data showed a positive correlation between the concentrations of bacteria and air humidity (p 0.002, R2 0.726) and between bacterial concentrations and age of the schools (p 0.045, R2 0.787). Also, that there was seasonal variation since the concentrations of fungi and bacteria varied according to the months (p 0.001).

Gabapentin reduces cardiovascular responses to laryngoscopy and tracheal intubation.

Background and objective: We have compared the effects of gabapentin on arterial pressure and heart rate at induction of anaesthesia and tracheal intubation in a randomized double‐blind study. Methods: Ninety normotensive patients (ASA I) undergoing elective surgery were divided into three groups of 30 patients each. Patients received oral placebo (Group I), 400 mg of gabapentin (Group II) or 800 mg of gabapentin (Group III) 1 h prior to surgery in the operating theatre. After induction of anaesthesia heart rate and mean arterial pressure were recorded at baseline 1, 3, 5, 10 and 15 min after intubation. Results: Patients receiving placebo and 400 mg gabapentin showed a significant increase in blood pressure and heart rate associated with tracheal intubation compared to baseline levels and Group III. There was significant decrease in heart rate and arterial pressure in Group III after intubation 1, 3, 5 and 10 min (P < 0.001, P < 0.001, P < 0.05 and P < 0.05, respectively) compared to Groups I and II. Conclusion: Given 1 h before operation gabapentin 800 mg blunted the arterial pressure and heart rate increase in first 10 min due to endotracheal intubation. Oral administration of gabapentin 800 mg before induction of anaesthesia is a simple and practical method for attenuating pressor response to laryngoscopy and tracheal intubation after standard elective induction.

Airborne Fungi and Actinomycetes Concentrations in the Air of Eskisehir City (Turkey)

The present study investigated the isolation and identification of airborne fungi from three different urban stations located in Eskisehir (Turkey). Air samples were taken by exposing a Petri dish with Rose-Bengal streptomycin agar medium for 15 min and after incubation the number of growing colonies was counted. The sampling procedure for fungi was performed 35 times at the research stations weekly between March and November 2001. A total of 2518 fungal and 465 actinomycetes colonies were counted on 420 Petri plates over a nine-month period. In total, some 20 mould species belonging to 12 genera were isolated. Alternaria alternata, Cladosporium cladosporioides and Scopulariopsis brevicaulis were the most abundant species in the study area (13.66, 5.80 and 5.50% of the total, respectively). Relationships between fungal spore numbers, aerosol air pollutants (that is the particulate matter in the air) and sulphur dioxide together with the meteorological conditions were examined using statistical analysis. Number of fungi and actinomycetes were tested by multivariate analysis (MANOVA) according to the areas and months. Fungal numbers were nonsignificant according to the areas and months (p > 0.05), but the number of actinomycetes recorded was significant (p < 0.01).

Prognostic value DCE-MRI parameters in predicting factor disease free survival and overall survival for breast cancer patients

PURPOSE The aim of the study is to assess the predictive power of DCE-MRI semi-quantitative parameters during treatment of breast cancer, for disease-free (DFS) and overall survival (OS). MATERIALS AND METHODS Forty-nine women (age range, 28-84 years; mean, 50.6 years) with breast cancer underwent dynamic contrast enhancement MRI at 1.0T imaging, using 2D FLASH sequences. Time intensity curves (TICs) were obtained from the regions showing maximal enhancement in subtraction images. Semi-quantitative parameters (TICs; maximal relative enhancement within the first minute, E (max/1); maximal relative enhancement of the entire study, E(max); steepest slope of the contrast enhancement curve; and time to peak enhancement) derived from the DCE-MRI data. These parameters were then compared with presence of recurrence or metastasis, DFS and OS by using Cox regression (proportional hazards model) analysis, linear discriminant analysis. RESULTS The results from of the 49 patients enrolled into the survival analysis demonstrated that traditional prognostic parameters (tumor size and nodal metastasis) and semi-quantitative parameters (E(max/1), and steepest slope) demonstrated significant differences in survival intervals (p<0.05). Further Cox regression (proportional hazards model) survival analysis revealed that semi-quantitative parameters contributed the greatest prediction of both DFS, OS in the resulting models (for E(max/1): p=0.013, hazard ratio 1.022; for stepest slope: p=0.004, hazard ratio 1.584). CONCLUSION This study shows that DCE-MRI has utility predicting survival analysis with breast cancer patients.

Effects of coffee consumption and smoking habit on bone mineral density.

This study aims to investigate how a person’s smoking and coffee consumption habits in the premenopausal stage can affect the postmenopausal BMD values. Two hundred females in the postmenopause stage were evaluated. The average daily coffee consumption and smoking habits in the premenopause stage and the demographic characteristics, age and duration of menopause of all the cases were identified and noted. The bone mineral density (BMD) evaluations of these cases were made with Dual Energy X-ray Absorbsiometer (DEXA) technique. The relationship of the questioned risk factors with BMD and differences among the groups were investigated. No correlation was found between the amount of coffee consumption and BMD. The BMD values of the smokers’ group were lower than non-smokers’ group. As a result, advancing age, duration of menopause and smoking habits have been identified to be risk factors in relation to OP.