Yazar "Eroglu, Yesim" seçeneğine göre listele

Listeleniyor 1 - 3 / 3
  • Küçük Resim Yok
    Öğe
    Classification of computerized tomography images to diagnose non-small cell lung cancer using a hybrid model
    (Springer, 2023) Demiroglu, Ugur; Senol, Bilal; Yildirim, Muhammed; Eroglu, Yesim
    Lung cancer arises from the abnormal and uncontrolled reproduction of parenchymal cells. Among all cancer cases, lung cancer is one of the prevailing types. Prevalence and death rates of the cancer increases day by day. From this point of view, early diagnosis and treatment of this cancer increases survival times and rates. The main idea in the development of the method presented in this publication is to increase the rate of early diagnosis. Computerized Tomography (CT) is the major screening method when encountered with suspicious symptoms. The cancer can be determined with CT and besides subtyping can be done. Diagnosing the disease with the human eye can sometimes lead to the emergence of deficiencies. This is one of the problems faced today. In this direction, the study in this paper presents a hybrid method to predict and diagnose the lung cancer from CT images to minimize potential human errors. Using the method, feature maps of the CT images of the dataset are obtained using the previously trained DarkNet-53 and DenseNet-201 deep model architectures. DarkNet-53 and DenseNet-201 architectures were chosen because they gave the best feature extraction results among 7 different architectures. The purpose of using the two architectures is to combine two high-performance models to create a hybrid classification method with high accuracy. Feature concatenation is applied to increase the diagnosis accuracy. To optimize the performance and computation cost of the proposed method, the Neighborhood Component Analysis (NCA) optimization method is used in determining and analysis of the features with more information. Therefore, features with less contribution in the accuracy are eliminated. Next, new feature maps are achieved by grading all features upon their weights and applying an elimination using a threshold value. The new feature maps are classified using Classical Machine Learning (CML) classifiers. Classification accuracies on DarkNet-53 architecture were calculated as 69.11% with SoftMax and 96.25% with Ensemble Classifiers and Nearest Neighbor Classifiers respectively. Similarly, accuracies on DenseNet-201 architecture were calculated as 68.29% with SoftMax and 97.39% with Ensemble Classifiers and Nearest Neighbor Classifiers respectively. With the proposed hybrid model, the Ensemble Classifiers reached the accuracy of 98.69% and the highest accuracy is achieved by using k-Nearest Neighbor Classifier (kNN) with the value of 98.86%. The results are supported with detailed illustrations.
  • Küçük Resim Yok
    Öğe
    Comparison of Computed Tomography-Based Artificial Intelligence Modeling and Magnetic Resonance Imaging in Diagnosis of Cholesteatoma
    (Aves, 2023) Eroglu, Orkun; Eroglu, Yesim; Yildirim, Muhammed; Karlidag, Turgut; Cinar, Ahmet; Akyigit, Abdulvahap; Yalcin, Sinasi
    BACKGROUND: In this study, we aimed to compare the success rates of computed tomography image-based artificial intelligence models and magnetic resonance imaging in the diagnosis of preoperative cholesteatoma. METHODS: The files of 75 patients who underwent tympanomastoid surgery with the diagnosis of chronic otitis media between January 2010 and January 2021 in our clinic were reviewed retrospectively. The patients were classified into the chronic otitis group without cholesteatoma (n = 34) and the chronic otitis group with cholesteatoma (n = 41) according to the presence of cholesteatoma at surgery. A dataset was created from the preoperative computed tomography images of the patients. In this dataset, the success rates of artificial intelligence in the diagnosis of cholesteatoma were determined by using the most frequently used artificial intelligence models in the literature. In addition, preoperative MRI were evaluated and the success rates were compared. RESULTS: Among the artificial intelligence architectures used in the paper, the lowest result was obtained in MobileNetV2 with an accuracy of 83.30%, while the highest result was obtained in DenseNet201 with an accuracy of 90.99%. In our paper, the specificity of preoperative magnetic resonance imaging in the diagnosis of cholesteatoma was 88.23% and the sensitivity was 87.80%. CONCLUSION: In this study, we showed that artificial intelligence can be used with similar reliability to magnetic resonance imaging in the diagnosis of cholesteatoma. This is the first study that, to our knowledge, compares magnetic resonance imaging with artificial intelligence models for the purpose of identifying preoperative cholesteatomas.
  • Küçük Resim Yok
    Öğe
    Diagnosis of periventricular leukomalacia in children with artificial intelligence-based models developed using brain magnetic resonance images
    (Springer London Ltd, 2023) Eroglu, Yesim; Yildirim, Muhammed; Cinar, Ahmet
    Periventricular leukomalacia is periventricular white matter damage that develops due to hypoxia and ischemia of the brain. It is one of the leading causes of neurological and developmental problems in children that will affect their future lives. Therefore, the correct diagnosis is important for giving the appropriate treatment. The main imaging method used in the diagnosis is magnetic resonance imaging (MRI). In this study, we evaluated the detectability of periventricular leukomalacia with artificial intelligence models in MRIs in children. In the study, two new artificial intelligence-based models are proposed to classify brain MRIs. The first proposed model consists of 19 layers, and this new model was more successful than previously trained deep models for classifying MRIs. In addition, the number of layers is lower than the models accepted in the literature. In the second model we proposed, the features were taken from our first model and optimized with the neighborhood component analysis (NCA) method, and then classified in the wide neural network. The accuracy values obtained in the models we have proposed are 94.62 and 98.92%, respectively. These accuracy values show that our proposed model is successful in classifying MRIs.