Browsing by Subject "Resolution"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
Item Open Access Data from: Performance of compressed sensing based image reconstruction for photoacoustic imaging(Bilkent University, 2022-08) Çelebi, Sadık ÇağatayItem Open Access The effects of feedback and training on the performance of probability forecasters(Elsevier, 1992) Benson, P. G.; Önkal D.An experiment examined the effects of outcome feedback and three types of performance feedback-calibration feedback, resolution feedback, and covariance feedback - on various aspects of the performance of probability forecasters. Subjects made 55 forecasts in each of four sessions, receiving feedback prior to making their forecasts in each of the last three sessions. The provision of calibration feedback was effective in improving both the calibration and overforecasting of probability forecasters, but the improvement was not gradual; it occurred in one step, between the second and third sessions. Simple outcome feedback had very little effect on forecasting performance. Neither resolution nor covariance feedback affected forecasters' performances much differently than outcome feedback. However, unlike outcome feedback, the provision of performance feedback caused subjects to manage their use of the probability scale. Subjects switched from two-digit probabilities to one-digit probabilities, and those receiving calibration and resolution feedback also reduced the number of different probabilities they used. © 1992.Item Open Access Performance of compressed sensing based image reconstruction for photoacoustic imaging(2022-08) Çelebi, Sadık ÇağatayPhotoacoustic Imaging (PAI) is an emerging imaging modality that provides high resolution and image contrast. PAI employs Delay-and-Sum (DAS) beamforming in which data on all array elements are processed for image reconstruction. In order to decrease system cost in PAI, it is desired to use fewer samples in image reconstruction. Reducing the number of active elements of the array without changing the aperture decreases the computational cost. However, spatial undersampling results in poor image resolution and contrast, and causes spatial artifacts called grating lobes. Compressed Sensing (CS) is a data completion scheme that alleviates the effects of undersampling using a priori knowledge of the signal of interest and the measurement scheme. We performed simulations and experiments to compare the performances of the CS image reconstruction algorithm and conventional DAS beamforming. We used Full-Width Half Maximum (FWHM) resolution and image contrast ratio (CR) as performance metrics. Simulation results offer improved image resolution and contrast when CS is used. Lateral resolution in DAS beamformed images deteriorates with depth. A lateral resolution of 150 µm is obtained regardless of depth using only a quarter of the transducer elements. However, the resolution of DAS beamformed images ranges between 255 µm to 508 µm as the depth increases. It is also shown that CS suppresses the effects of grating lobes and improves the contrast ratio up to 14 dB. We also presented the experimental verification of the results. We used an ultrasound research scanner, a tunable laser system, and an optoacoustic phantom in the experiments. We experimentally showed that the CS method mitigates the effects of spatial undersampling and outperforms the DAS beamforming method in terms of contrast by 10.81 dB on average. CS method also offers an improved lateral resolution of approximately 350 µm compared to 750 µm in DAS beamforming.Item Open Access Super-resolution diffusion model for accelerated MRI reconstruction(IEEE - Institute of Electrical and Electronics Engineers, 2023-08-28) Mirza, Muhammad Usama; Çukur, TolgaMRI reconstruction is a process to generate high-quality images from the raw data obtained during magnetic resonance imaging. Diffusion models, a class of generative models, have become a popular method for MRI Reconstruction due to their ability to generate high quality images. Diffusion models work by adding Gaussian noise to the original image and training a network to remove the noise. Diffusion models can continue to generate high quality images even with a different type of noise added to the original image. In this study we combine a resolution decreasing operator with noise scheduling used by regular diffusion models, ResDiff to perform MRI Reconstruction. One of the biggest drawbacks of Diffusion models is the amount of time taken to generate images. Down-sampling images to a lower resolution requires fewer steps allowing ResDiff to achieve competitive results in far less time.