Yazar "Eliiyi, Ugur" seçeneğine göre listele

Listeleniyor 1 - 3 / 3
  • Küçük Resim
    Öğe
    Deep Learning-based Classification of Breast Tumors using Raw Microwave Imaging Data
    (Gazi Univ, 2024) Bicer, Mustafa Berkan; Eliiyi, Ugur; Tursel Eliiyi, Deniz
    Breast cancer is the leading type of malignant neoplasm disease among women worldwide. Breast screening makes extensive use of powerful techniques such as x-ray mammography, magnetic resonance imaging, and ultrasonography. While these technologies have numerous benefits, certain drawbacks such as the use of low-energy ionizing x-rays, a lack of specificity for malignant tissues, and cost, have motivated researchers to investigate novel imaging and detection modalities. Microwave imaging (MWI) has been extensively studied due to its low-cost structure and ability to perform measurements using non-ionizing electromagnetic waves. This study proposes a novel convolutional neural network (CNN) model for detecting and classifying tumor scatterers in MWI simulation data. To accomplish this, 10001 different numerical breast models with tumor scatterers of varying numbers and positions were developed, and the simulation results were derived using the synthetic aperture radar (SAR) technique. The presented CNN structure was trained using 8000 pieces of simulation data, and the remaining data were used for testing, achieving accuracy rates of 99.61% and 99.75%, respectively. The proposed model is compared to three state-of-the-art models on the same dataset in terms of classification performance. The results demonstrate that the proposed model effectively performs effectively well in detecting and classifying tumor scatterers.
  • Küçük Resim
    Öğe
    Detection of Freezing of Gait Episodes in Patients with Parkinson's Disease using Electroencephalography and Motion Sensors: A Protocol and its Feasibility Results
    (Wolters Kluwer Medknow Publications, 2022) Eliiyi, Ugur; Kahraman, Turhan; Genc, Arzu; Keskinoglu, Pembe; Ozkurt, Ahmet; Donmez, BerrilColakoglu
    Objective: Freezing of gait (FOG) is an important concern for both patients with Parkinson's disease (pwPD) and physicians. In this study, we aimed to introduce a study protocol and our initial data. The data were subsequently used in machine learning models to detect FOG episodes using brain activity signals and motion data in the laboratory setting using complex FOG-evoking activities in a sample of pwPD with and without FOG compared with age-matched healthy controls. Subjects and Methods: An experimental task to evoke a FOG episode was designed. This experimental task was tested on two pwPD with FOG in on and off periods and one healthy control. Brain activity signals and motion data were collected simultaneously using electroencephalography (EEG) and inertial measurement units (IMUs). Results: The whole procedure took about 2 h, during which around 30 min were spent on walking tasks, involving 35 complete tours in the designed 8-m hallway by pwPD. Both EEG and IMUs sensor data could be collected, accompanied by FOG episode data marked by the neurologist. The video recordings of the patient's walking tasks were checked and reanalyzed by the neurologist sometime after the data experiment for marking the beginnings and ends of the observed FOG episodes more precisely. In the end, 24 stops were marked as FOG, which corresponded to 11% of the sensor data collected during the walking tasks. Conclusion: The designed FOG-evoking task protocol could be performed without any adverse effects, and it created enough FOG episodes for analysis. EEG and motion sensor data could be successfully collected without any significant artifacts.
  • Küçük Resim
    Öğe
    A DYNAMIC PACKING APPROACH FOR INTERNET-OF-THINGS AND LOGISTICS APPLICATIONS
    (Inst Applied Mathematics, 2023) Eliiyi, Ugur; Nasibov, Efendi
    A novel dynamic rectangular packing model for a resource allocation problem, which has many applications in Logistics Internet-of-Things (L-IoT) is considered in this study. We use an optimization approach to deal with an IoT-based problem, whose objective is to max-imize the profit obtained from packing the data demand over a sequence of time frames, while satisfying several quality of service constraints. We propose a nonlinear integer programming model for the problem that optimizes demand partitioning and rectangle packing simultaneously, for the first time in literature. By introducing effective upper and lower bounds for this practi-cally important problem, a computational experimentation is designed for assessing the model and bound performances. An extensive discussion and recommendations for policy-making are included based on the computational results.