A5071792338- Medical Image Segmentation Techniques
- Radiomics and Machine Learning in Medical Imaging
- Advanced Neural Network Applications
- COVID-19 diagnosis using AI
- Geophysical Methods and Applications
- Computer Graphics and Visualization Techniques
- Video Surveillance and Tracking Methods
- Medical Imaging Techniques and Applications
- Industrial Vision Systems and Defect Detection
- AI in cancer detection
- Advanced Data Compression Techniques
- Ultrasonics and Acoustic Wave Propagation
- Autonomous Vehicle Technology and Safety
- Advanced X-ray and CT Imaging
- Medical Imaging and Analysis
- Functional Brain Connectivity Studies
- Brain Tumor Detection and Classification
- EEG and Brain-Computer Interfaces
- Acute Ischemic Stroke Management
- Advanced SAR Imaging Techniques
- Underwater Acoustics Research
- Mineral Processing and Grinding
- Image and Object Detection Techniques
- Electromagnetic Scattering and Analysis
- Cerebrovascular and Carotid Artery Diseases
Dokuz Eylül University
2016-2025
Technologies pour la Santé
2024
Sivas State Hospital
2020
Etlik Zübeyde Hanım Kadın Hastalıkları Eğitim ve Araştırma Hastanesi
2018
To compare the accuracy and repeatability of emerging machine learning based (i.e. deep) automatic segmentation algorithms with those well-established semi-automatic (interactive) methods for determining liver volume in living transplant donors at computerized tomography (CT) imaging.
AbstractThe inspection of the fabric defects is an important problem, which highly affects both quality and cost in textile industry. Because consistency accuracy problems, defect by human experts neither feasible nor efficient. This requires development use automated techniques. Thus, this study, a texture analysis method, uses sum difference histograms (SDH) conjointly with co-occurrence matrices, proposed to introduce objective criterion for detection. To accomplish detection task high...
This study uses machine learning (ML) to elucidate the contact relationship between mandibular third molar (M3M) and inferior alveolar canal (IAC), leading three major contributions; (1) The first publicly accessible PR image dataset with semantic annotations for 1,478 IACs M3Ms from 1,010 patients is introduced, which includes challenging cases, such as false positive contacts, CBCT images gold standard, (2) Established radiological indicators M3M-IAC were extracted features using digital...
Diagnosing bipolar disorder (BD) and schizophrenia (SCH) presents significant challenges due to overlapping symptoms, reliance on subjective assessments, the late-stage manifestation of many symptoms. Current methods using structural magnetic resonance imaging (sMRI) as input data often fail provide objectivity sensitivity needed for early accurate diagnosis. sMRI is well known be capable detecting anatomical changes, such reduced gray matter volume in SCH or cortical thickness alterations...
To evaluate the effectiveness of machine learning (ML) models in predicting occurrence retinopathy prematurity (ROP) and treatment need. Four ML were created using 49 parameters known within first 24 h post-birth obtained during initial screening examination, encompassing demographic, maternal, clinical, neonatal intensive care unit-related data. The models' performances assessed five classifier algorithms: logistic regression (LR), decision tree (DT), support vector (SVM), random forest...
Background: The vitreomacular interface (VMI) encompasses a group of retinal disorders that significantly impact vision, requiring accurate classification for effective management. This study aims to compare the effectiveness an expert-designed custom deep learning (DL) model and code free Auto Machine Learning (ML) in classifying optical coherence tomography (OCT) images VMI disorders. Materials Methods: A balanced dataset OCT across five classes—normal, epiretinal membrane (ERM),...
As being a tool that assigns optical parameters used in interactive visualization, Transfer Functions (TF) have important effects on the quality of volume rendered medical images. Unfortunately, finding accurate TFs is tedious and time consuming task because trade off between using extensive search spaces fulfilling physician's expectations with data exploration tools interfaces. By addressing this problem, we introduce semi-automatic method for initial generation TFs. The proposed uses Self...
Determining the position and alignment of rails ahead a train is an essential element camera based driver support systems, which aim to detect obstacles around railway. Effective extraction through faces many challenges that cause poor visibility due several factors including but not limited bad weather conditions, inverse illumination, shadows, rust, clustered rails. Moreover, process should be completed almost in real time without using any explicit knowledge about speed or parameters....
Marble quality classification is an important procedure generally performed by human experts. However, using experts for error prone and subjective. Therefore, automatic computerized methods are needed in order to obtain reproducible objective results. Although several proposed this purpose, we demonstrate that their performance limited when dealing with diverse datasets containing a large number of groups. In work, test feature sets neural network topologies better performance. During these...
Pedestrian Detection (PD) is one of the most studied issues driver assistance systems. Although a tremendous effort already given to create datasets and develop classifiers for cars, studies about railway systems remain very limited. This article shows that direct application neither existing advanced object detectors (such as AlexNet, VGG, YOLO etc.), nor specifically created PD Caltech/INRIA trained classifiers), can provide enough performance overcome specific challenges. Fortunately, it...
Classification of objects from scattered electromagnetic waves is a difficult problem, as it heavily depends on aspect angle. To minimize this dependency, distinguishable features can be used. In paper, we propose target identification method in the resonance scattering region using novel structural feature set based signal waveform. obtain robustness at low signal-to-noise ratio (SNR), multiscale approximation used for distortion correction prior to extraction. This achieved by an...
Intuitive and differentiating domains for transfer function (TF) specification direct volume rendering is an important research area producing informative useful 3D images. One of the emerging branches this texture based functions. Although several studies in two, three, four dimensional image processing show importance using information, these generally focus on segmentation. However, TFs can also be built effectively appropriate information. To accomplish this, methods should developed to...
In medical visualization, segmentation is an important step prior to rendering. However, it also a difficult procedure because of the restrictions imposed by variations in image characteristics, human anatomy, and pathology. Moreover, what interesting from clinical point view usually not only organ or tissue itself, but its properties together with adjacent organs related vessel systems that are going coming out. For informative rendering, these necessitate usage different methods single...
Tomographic medical imaging systems produce hundreds to thousands of slices, enabling three-dimensional (3D) analysis. Radiologists process these images through various tools and techniques in order generate 3D renderings for applications, such as surgical planning, education, volumetric measurements. To save store visualizations, current use snapshots or video exporting, which prevents further optimizations requires the storage significant additional data. The Grayscale Softcopy...