Yazar "Yilmaz, Hakan" seçeneğine göre listele
Listeleniyor 1 - 11 / 11
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe FahamecV1:A Low Cost Automated Metaphase Detection System(Eos Assoc, 2017) Yilmaz, Hakan; Turan, M. KamilIn this study, FahamecV1 is introduced and investigated as a low cost and high accuracy solution for metaphase detection. Chromosome analysis is performed at the metaphase stage and high accuracy and automated detection of the metaphase stage plays an active role in decreasing analysis time. FahamecV1 includes an optic microscope, a motorized microscope stage, an electronic control unit, a camera, a computer and a software application. Printing components of the motorized microscope stage (using a 3D printer) is of the main reasons for cost reduction. Operations such as stepper motor calibration, are detection, focusing, scanning, metaphase detection and saving of coordinates into a database are automatically performed. To detect metaphases, a filter named Metafilter is developed and applied. Average scanning time per preparate is 77 sec/cm(2). True positive rate is calculated as 95.1%, true negative rate is calculated as 99.0% and accuracy is calculated as 98.8%.Öğe Filter Development for Automatic Detection of Analyzable Metaphases(Ieee, 2018) Yilmaz, Hakan; Turan, M. KamilAbnormalities in the structure of chromosomes cause fetal deaths or developmental disorders. Chromosome analysis is a method used to diagnose many chromosomal disorders such as Down syndrome. Metaphase images are needed for chromosome analysis. Objective selections must be made during the acquisition of these images. Selecting of non-analyzable images could directly affect the results of chromosome analysis. In this study, a filter was developed that automatically detects analyzable metaphase images. The developed filter was used with the motorized microscope table and the analyzable metaphase images were detected. After expert evaluation on the results obtained, the average success rate of the filter was calculated as 98.9%. The filter performed an average run time of 76 milliseconds per square.Öğe Multiple classification of EEG signals and epileptic seizure diagnosis with combined deep learning(Elsevier, 2023) Varli, Muhammet; Yilmaz, HakanEpilepsy stands out as one of the common neurological diseases. The neural activity of the brain is observed using electroencephalography (EEG), which allows the diagnosis of epilepsy disease. The aim of this study is to create a combined deep learning model that automatically detects epileptic seizure activity, detection of the epileptic region and classifies EEG signals by using images representing the time-frequency components of the time series EEG signal and numerical values of the raw EEG signals. In the study, 3 different public datasets, CHB-MIT, BernBarcelona and Bonn EEG records were used. This study presents a combined model using the time sequence of EEG signals and time-frequency-image transformations of time-dependent EEG signals. CWT and STFT methods were used to convert signals to images. Two models were created separately with the images created by CWT and STFT methods. In the Bonn dataset average accuracy rates of 99.07 %, 99.28 %, respectively, in binary classifications and 97.60 % and 98.56 %, respectively, in multiple classifications were obtained with scalogram and spectrogram images. In the Bern-Barcelona and CHB-MIT datasets, 95.46 % and 96.23 % accuracy rates were obtained, respectively. The data combinations brought together in 3 different combinations with the Bonn dataset were underwent to 8-fold cross validation and average accuracy rates of 99.21 % (+/- 0.56), 99.50 % (+/- 0.45), and 98.84 % (+/- 1.58) were obtained. The model we created can detect whether there is epileptic seizure activity in EEG data, detection of the epileptic region and classify EEG signals with a high success rate.Öğe Non-invasive hemoglobin estimation from conjunctival images using deep learning(Elsevier Sci Ltd, 2023) Cuvadar, Beyza; Yilmaz, HakanHemoglobin, a crucial protein found in erythrocytes, transports oxygen throughout the body. Deviations from optimal hemoglobin levels in the blood are linked to medical conditions, serving as diagnostic markers for certain diseases. The hemoglobin level is usually measured invasively with different devices using the blood sample. In the physical interpretation, some signs are traditionally used. These signs are the palms, face, nail beds, pallor of the conjunctiva, and palmar wrinkles. Studies have shown that conjunctival pallor can yield more effective results in detecting anemia than the pallor of the palms or nail beds.This study is aimed to predict the hemoglobin level by deep learning method, non-invasive, cheap, fast, high accuracy, and without creating medical waste. In this context, conjunctival images and age, weight, height, gender, and hemoglobin values were collected from 388 people who donated blood to the Turkish Red Crescent. A dataset was generated by augmenting the gathered data with body mass index data. Within the scope of this investigation, the limits of agreement (LoA) value at a 95% confidence interval was computed to be 1.23 g/dL, while the bias was established as 0.26 g/dL. The mean absolute percentage error (MAPE) values were determined to be 3.4%, and the root mean squared error (RMSE) was calculated to be 0.68 g/dL. These findings exhibit a successful outcome compared to similar investigations, signifying that this non-invasive method can be employed for hemoglobin level estimation. Furthermore, the estimated hemoglobin levels could aid in diagnosing several hemoglobin-related ailments.Öğe A novel combined deep learning methodology to non-invasively estimate hemoglobin levels in blood with high accuracy(Elsevier Sci Ltd, 2022) Yilmaz, Hakan; Kizilates, Burcu S.; Shaaban, Fatema; Karatas, Ziya R.Hemoglobin is an essential protein found in blood and should not fall below a certain level in humans. Today's methods of hemoglobin measurement are mostly invasive. This study aims to perform a non-invasive estimation of hemoglobin levels using age, height, weight, body mass index, gender, and nail images of individuals. Data was collected from 353 volunteers aged 1 to 92 years. Two different data sets were created using these data: a numerical dataset and a nail image set. A combined deep learning model was put forward using both the model created for numerical data and the model created for nail images. In this study, bias was calculated as 0.03 g/dL, and the limits of agreement value in the 95% confidence interval was calculated as 1.09 g/dL. The calculated mean absolute percentage error values were 2.09%, and the root mean squared error was 0.56 g/dL. After entering the necessary data into the system, the estimated average resulting time was 0.09 s. The results of this study have shown success compared to the results of similar studies, and this method can be used for non-invasive hemoglobin level estimation. The recommended approach is more comfortable than invasive methods and gives much faster results.Öğe Optimization of 3D Printing Operation Parameters for Tensile Strength in PLA Based Sample(Gazi Univ, 2020) Gunay, Mustafa; Gunduz, Suleyman; Yilmaz, Hakan; Yasar, Nafiz; Kacar, RamazanIn this study, the mechanical properties of PLA+ samples produced by using fused deposition method (FDM) based 3D printer were investigated in detail for the effects of printing speed, infill rate and raster angle. For this purpose, standard tensile test specimens were prepared with a 3D printer according to Taguchi L-18 experimental design. The effects on the tensile strength of the process parameters (printing speed, infill rate and raster angle) were determined by analysis of variance (ANOVA). In addition, the process parameters for the tensile strength were optimized by applying the Taguchi methodology. Consequently, while the most effective parameter on the tensile strength was the infill rate, the raster angle and the printing speed were determined as other important parameters, respectively.Öğe Prediction and classification of tool wear and its state in sustainable machining of Bohler steel with different machine learning models(Elsevier Sci Ltd, 2023) Korkmaz, Mehmet Erdi; Gupta, Munish Kumar; Kuntoglu, Mustafa; Patange, Abhishek D.; Ross, Nimel Sworna; Yilmaz, Hakan; Chauhan, SumikaMachine learning has numerous advantages, especially in the rapid digitization of the manufacturing industry that combines data from manufacturing processes and quality measures. Predictive quality allows manufacturers to make informed predictions about the quality of their products by analyzing data gathered during production. The quality of the machining, the total cost and the computation time need to be improved using contemporary production processes. With this concern, a series of experiments were carried out on Bohler steel both in dry, Minimum Quantity Lubrication (MQL) and nano-MQL conditions in varying quantities to explore the tool wear. In comparison to dry conditions, the utilization of MQL in machining processes demonstrates significantly enhanced efficacy in mitigating flank wear. The reduction in flank wear ranges from around 5% to 20% to 25%, contingent upon the application of MQL on the flank face, rake face, or both faces simultaneously. After that, the results of the tests were evaluated with the models of machine learning (ML) to determine which environment was optimal for cutting under both real and artificial circumstances.Öğe Prediction of gastric cancer by machine learning integrated with mass spectrometry-based N-glycomics(Royal Soc Chemistry, 2023) Demirhan, Deniz Baran; Yilmaz, Hakan; Erol, Harun; Kayili, Haci Mehmet; Salih, BekirEarly and accurate diagnosis of gastric cancer is vital for effective and targeted treatment. It is known that glycosylation profiles differ in the cancer tissue development process. This study aimed to profile the N-glycans in gastric cancer tissues to predict gastric cancer using machine learning algorithms. The (glyco-) proteins of formalin-fixed parafilm embedded (FFPE) gastric cancer and adjacent control tissues were extracted by chloroform/methanol extraction after the conventional deparaffinization step. The N-glycans were released and labeled with a 2-amino benzoic (2-AA) tag. The MALDI-MS analysis of the 2-AA labeled N-glycans was performed in negative ionization mode, and fifty-nine N-glycan structures were determined. The relative and analyte areas of the detected N-glycans were extracted from the obtained data. Statistical analyses identified significant expression levels of 14 different N-glycans in gastric cancer tissues. The data were separated based on the physical characteristics of N-glycans and used to test in machine-learning models. It was determined that the multilayer perceptron (MLP) was the most appropriate model with the highest sensitivity, specificity, accuracy, Matthews correlation coefficient, and f1 scores for each dataset. The highest accuracy score (96.0 +/- 1.3) was obtained from the whole N-glycans relative area dataset, and the AUC value was determined as 0.98. It was concluded that gastric cancer tissues could be distinguished from adjacent control tissues with high accuracy using mass spectrometry-based N-glycomic data.Öğe Prediction of power consumption and its signals in sustainable turning of PH13-8Mo steel with different machine learning models(Springer London Ltd, 2024) Yurtkuran, Hakan; Korkmaz, Mehmet Erdi; Gupta, Munish Kumar; Yilmaz, Hakan; Gunay, Mustafa; Vashishtha, GovindDue to extensive distribution and huge demand of energy efficient processes, the energy-saving of machining processes draws more and more attention, and a significant variety of methods have evolved to prognosis or optimise the energy consumption in machining operations. Similarly, the estimation of power consumption-cutting conditions relationships is of great importance for optimizing processing costs and for cleaner machining. Compared to traditional methods, machine learning (ML) approach is one of the effective analysis options to model machinability indicators such as cutting force, tool wear, power consumption and surface quality. In this study, PH13-8Mo stainless steel was machined with coated carbide inserts using primarily Dry, MQL, nano-Graphene + MQL, nano-hBN + MQL, Cryo, Cryo + MQL cutting environments. Power consumption and its signals during milling were measured and different machine learning models were applied to estimate the Pc. To develop the Pc model based on the ML algorithm, 70% of the power consumption data is reserved for training and 30% for testing. In all cutting environments, power consumption increased by an average of 3.14% as feed speed increased. The reduction in Pc compared to the dry cutting was calculated as an average of 2.2%, 3.17%, 2.57%, 4.88% and 5.45% for MQL, nano-Graphen + MQL, nano-hBN + MQL, Cryo, Cryo + MQL, respectively. It is seen that the developed prediction model can reflect the power consumption-parameter relationships at high accuracy.Öğe Real-time monitoring and measurement of energy characteristics in sustainable machining of titanium alloys(Elsevier Sci Ltd, 2024) Gupta, Munish Kumar; Korkmaz, Mehmet Erdi; Yilmaz, Hakan; Sirin, Senol; Ross, Nimel Sworna; Jamil, Muhammad; Krolczyk, Grzegorz M.The development of cutting-edge monitoring technologies such as embedded devices and sensors has become necessary to ensure an industrial intelligence in modern manufacturing by recording machine, process, tool, and energy consumption conditions. Similarly, machine learning based real time systems are popular in the context of Industry 4.0 and are generally used for predicting energy needs and improving energy utilization efficiency and performance. In addition, sustainable and energy-efficient machining technologies that can reduce energy consumption and associated negative environmental effects have been the latest topic of much study in recent years. Concerning this regard, the present work firstly deals with the real time monitoring and measurement of energy characteristics while machining titanium alloys under dry, minimum quantity lubrication (MQL), liquid nitrogen (LN2) and hybrid (MQL + LN2) conditions. The energy characteristics at different stages of machine tools were monitored with the help of a high end energy analyser. Then, the energy signals from each stage of machining operation were predicted and classified with the help of different machine learning (ML) models. The experimental results showed that MQL, LN2, and hybrid conditions decreased the total energy consumption by averagely 2.6 %, 17.0 %, and 16.3 %, respectively, compared to dry condition. The ML results demonstrated that the accuracy of the random forest (RF) approach obtained higher efficacy with 96.3 % in all four conditions. In addition, it has been noticed that the hybrid cooling conditions are helpful in reducing the energy consumption values at different stages.Öğe Towards specific cutting energy analysis in the machining of Inconel 601 alloy under sustainable cooling conditions(Elsevier, 2023) Korkmaz, Mehmet Erdi; Gupta, Munish Kumar; Yilmaz, Hakan; Ross, Nimel Sworna; Boy, Mehmet; Sivalingam, Vinoth Kumar; Chan, Choon KitCurrently, the research efforts on machining indices such as tool wear, surface roughness, power consumption etc. is well reported in literature, but energy analysis based on material removal methods and machine learning has received comparatively little attention. Therefore, the present work deals with the research efforts on simultaneous reduction of specific cutting energy in sustainable machining of Inconel 601 alloy with different machine learning models. The studies were conducted using dry, minimum quantity lubrication (MQL), nano-MQL, cryogenic, and hybrid cooling methods (cryo-nano-MQL). The specific cutting energy (SCE) values were calculated based on the data obtained from power consumption and material removal rate. Subsequently, the SCE data is employed to construct the crucial maps, which are then utilized in several sophisticated machine learning models, including Multiple Linear Regression, Lasso Regression, Bayesian Ridge Regression, and Voting Regressor, to facilitate the predictive modeling of outcomes. The findings of the study indicate that the Bayesian model exhibits a comparatively reduced error rate and a closely aligned R2 value when compared to other prediction models. Moreover, as a novelty, nanoparticles addition into hybrid cooling methods (cryo + nano + MQL) also showed better performance as well as 0.3 % less specific cutting energy than only cryo method which is previously used in former studies.
