Research

Machine Learning - AI

In Determination of Fiber Tracts on the Diffusion Phantom Using Q-Routing Method, the application of Q-routing method to the diffusion MR fiber tractography problem is proposed. The results reveal the potential of the proposed method on the diffusion MR fiber tractography problem. Paper
In the IronTract Challenge, the accuracy was evaluated by comparison to tracer injections in the same macaque brains as the diffusion data. The accuracy of the proposed method (5th method) was comparable with the state of the art methods. ISMRM Abstract		 Image
In Landmark Localization On Color Coded Diffusion Anisotropy Images Using Convolutional Neural Networks, a CNN based landmark detector is employed to locate specific landmarks at given MNI coordinates, on an individual’s color FA images of the brain. Although, the CNN training is performed by using the FNIRT labelled landmarks, the performance of CNN is higher at some regions. Therefore, the CNN does not learn the FNIRT labelling process but the underlying diffusion pattern. OHBM Poster Image
The aim of Short-Term Forecasting COVID-19 Cases In Turkey Using Long Short-Term Memory Network is to evaluate the value of Long Short-Term Memory (LSTM) Networks in forecasting the total number of COVID-19 cases in Turkey. Holt-Winters Additive method with Damped Trend gives superior results to LSTM Networks in forecasting the total number of COVID-19 cases. Addition of the number of deaths data to the input, results a lower error in forecasting, compared to using only the number of total cases as the input. Paper Image
Image In Tooth Instance Segmentation on Panoramic Dental Radiographs Using U-Nets and Morphological Processing we propose a post-processing stage to obtain a segmentation map in which the objects in the image are separated, and apply this technique to tooth instance segmentation with U-Net network. Paper HF Demo
Measurement of Merchant Angles is a fully automated method to measure Patellar Congruence Angle and Tilt Angle on Merchant Knee radiographs, employing CNN landmark localizers. Image

Diffusion MR Imaging

Cellular Automata Tractography method, which runs on GPU, allows shortest-paths tractography under one second runtime. Furthermore, connectivity based segmentation of the cerebral structures can be performed at the same time, as shown in the figure. Code and Article
In the IronTract Challenge, the accuracy was evaluated by comparison to tracer injections in the same macaque brains as the diffusion data. The accuracy of the proposed method (5th method) was between the deterministic and probabilistic streamline methods. ISMRM Abstract		 Image
Image Diffusion tensor imaging in early amyotrophic lateral sclerosis study aimed to perform diffusion tensor imaging (DTI) to detect white matter microstructural abnormalities, and also understanding the pathophysiology in Amyotrophic lateral sclerosis (ALS) using 3T magnetic resonance imaging. DTI showed decreased fractional anisotropy in bilateral corticospinal tracts and internal capsule posterior crus. There was a correlation between the FA reductions in this region and the severity of the disease in the patients with ALS. Consequently, DTI may be useful to monitor the progress and effectiveness of treatment interventions, as well as understanding the pathophysiology of ALS. Article
Image Comparison of Inter-Regional and Inter-Subject Variations of Cortical NODDI and DTI Parameters, aims to investigate the relationship between the regional microstructural differences of the cerebral cortex and diffusion MR parameters, that are obtained by fitting a single tensor model or three compartmental NODDI model. The parameters measured by diffusion MRI have some regional specificities through the cortex. The discriminant function coefficients for each parameter indicates that OD of NODDI model is the most important parameter where ISOVF is less important to capture the regional structural differences of the cortex. OHBM Abstract

Functional MRI Applications

Effects of Menstrual Cycle on Resting State Functional Connectivity in Schizophrenia study investigates resting state functional connectivity (RS-FC) alterations related to menstrual cycle phase and/or hormone levels in patients with schizophrenia and healthy controls and aims to contribute to the understanding of the effects of ovarian hormones on the pathogenesis of schizophrenia and brain functions. Abstract

In Working memory dysfunction in delusional disorders, using an established working memory paradigm with variable levels of memory load, we investigated alterations in functional magnetic resonance imaging (fMRI) of brain regions in patients with Delusional disorder (DD). Our findings were in line with the findings of the early reports on deficient functioning in temporal or limbic regions of the brain. Further, patients with DD displayed prefrontal dysfunction as seen in patients with schizophrenia. Article Image

Functional MRI in Determining Language Lateralization aims to noninvasively measure the laterality of language functions in the brain for the presurgical evaluation of epileptic patients. Questions for the word production task were projected to a screen and were seen by the patient through a mirror centered above the subjects eyes. The results of this f-MRI study have shown that a word generation task is reliable for determining the hemispheric language dominance. Article

Student Projects

Image Ultrasonography is the method that is routinely applied for screening and classification of hip dysplasia in newborns. Phantoms, that can mimic the acoustic characteristics of human tissues, are needed in ultrasound training or in the calibration of devices. In Newborn Hip Ultrasonography Training Phantom polylactic acid (PLA) for simulating the bone tissue and gelatin for soft tissue were aimed to create a newborn hip ultrasonography phantom for training purposes. Paper
Cranioplasty is the surgical operation to repair the cranial defects both in cosmetic and functional ways. In 3D Printing and PMMA Molding Study for a Cranioplasty Application, a method to copy the geometry of the cranial bone model constructed in digital environment to PMMA by using 3d printing and molding technique, is presented. The factors, affecting the error, are evaluated by comparing the geometries arise at each step of the multi-stage technique. Paper Image
Image Epileptic Focus Identification Based on Hemispheric Asymmetry using Surface Based Analysis study aimed to localize the cortical malformations of epileptic patients using surface based morphological analysis. Paper