Browsing by Author "Arslan, Fuat"

Now showing 1 - 5 of 5

Restricted
Bir beldeyi ayağa kaldıran emektar santral: Tunçbilek Termik Santrali
(Bilkent University, 2020) Altınel, Mustafa Salih; Tekin, İpek; Arslan, Fuat; Akyıldız, Eda; Kılkış, Öykü
1950’li yıllarda Türkiye’nin sürekli artan enerji ihtiyacını karşılaması gerekiyordu. 1950 yılında iktidara gelen Demokrat Parti söz konusu enerji ihtiyacına önem vererek enerji politikalarını bu alanda sürdürerek termik santrallerin kurulmasını kararlaştırdı. 1956 yılında kurulan Tunçbilek Termik Santrali bu enerji politikaları sonucunda açılmış santrallerden birisidir. Tunçbilek Termik Santrali aktif olduğu dönemde bölge halkının geçim kaynağı oldu ve aynı zamanda Orta Anadolu’nun elektrik ihtiyacını karşıladı. 1980’li yılların başlangıcında santralin çevreye verdiği zararların fark edilmesi ve santralin özelleştirilmesi ve makineleşme sonucunda Tunçbilek Termik Santrali halkın yaşamındaki yerini kaybetti ve 2020 yılının Ocak ayında kapandı.
Open Access
Generalizable deep mri reconstruction with cross-site data synthesis
(IEEE, 2024-06-23) Nezhad, Valiyeh Ansarian; Elmas, Gökberk; Arslan, Fuat; Kabas, Bilal; Çukur, Tolga
Deep learning techniques have enabled leaps in MRI reconstruction from undersampled acquisitions. While they yields high performance when tested on data from sites that the training data originates, they suffer from performance losses when tested on separate sites. In this work, we proposed a novel learning technique to improve generalization in deep MRI reconstruction. The proposed method employs cross-site data synthesis to benefit from multi-site data without introducing patient privacy risks. First, MRI priors are captured via generative adversarial models trained at each site independently. These priors are shared across sites, and then used to synthesize data from multiple sites. Afterwards, MRI reconstruction models are trained using these synthetic data. Experiments indicate that the proposed method attains higher generalization against single-site models, and higher site-specific performance against site-average models.
Open Access
Multi-contrast mr image synthesis with a brownian diffusion model
(IEEE, 2024-12-05) Kabaş, Bilal; Arslan, Fuat; Nezhad, Valiyeh Ansarian; Çukur, Tolga
Magnetic Resonance Imaging (MRI) plays a significant role in medical diagnostics. However, prolonged scan times may hinder its widespread applicability in clinical settings. To mitigate this challenge, certain contrasts within multi-contrast MRI protocols can be excluded, and these target contrasts can then be synthesized from the acquired set of source contrasts retrospectively. Recently introduced generative adversarial and diffusion based MRI synthesis models yield enhanced performance against classical methods, yet there can still benefit from technical improvements. In this study, we propose a Brownian diffusion-based multi-contrast MR image synthesis model. Existing diffusion models synthesize images starting from a Gaussian noise sample, so guidance from the source contrast images are weakened. Conditional denoising diffusion models employs a weak conditioning during reverse process within the denoising network that may result in suboptimal sample generation due to poor convergence to target distribution. Capitalizing Brownian diffusion, the proposed model instead incorporates stronger guidance toward the target contrast distribution via a refined diffusion process. Experimental results suggest that our method attains higher performance in noise reduction and capture of tissue structural details over existing methods.
Open Access
Robust brain tumor segmentation with deep residual supervision and mixed precision training
(IEEE, 2024-06-23) Arslan, Fuat; Yılmaz, Melih Berk; Çukur, Tolga
Segmentation of brain tumors from MRI data is an application of great clinical importance in diagnostic evaluation, treatment and operational planning processes. In recently proposed deep learning techniques, supervision is commonly applied to network output segmentation maps, which may lead to deficiencies in learning features in early network stages. In addition, early termination of training or restricting the number of model parameters in order to limit the computational load caused by three-dimensional architectures that process volumetric MRI data may cause performance losses. The novel segmentation method proposed in this study enhanced sensitivity to information in MR images by applying deep residual supervision on feature maps in decoder stages of the neural network. Additionally, it reduces computational complexity by using mixed precision training algorithms, thus providing effective training in short run times. Experiments on the BraTS dataset show that the proposed model yields higher performance than reference techniques while improving computational efficiency.
Open Access
Super resolution mri via upscaling diffusion bridges
(2024-06-23) Mirza, Muhammad Usama; Arslan, Fuat; Çukur, Tolga
Magnetic Resonance Imaging (MRI) is a powerful medical imaging modality that provides high-resolution anatomical information about tissues. However, the intrinsic trade-off between acquisition time and image quality poses challenges in obtaining high-resolution images within a clinically feasible timeframe. This study introduces a novel approach to acquire high-resolution images in short scan times based on Super-Resolution Diffusion Bridges (SRDB). The proposed method leverages advanced machine learning techniques based on diffusion models to upscale MR images. The While standard diffusion models learn a mapping from Gausssian distributed noise images to target images, SRDB instead learns a mapping from low-resolution MR images to high-resolution images. Unlike the task-independent learning in standard diffusion model, SRDB thus performs task-based learning to improve structural consistency and better preservation of anatomical features. In this way, the trained models help capture fine details that may be missed in standard low-resolution MRI acquisitions.