Quality Enhancement of Computed Tomography Images of Porous Media Using Convolutional Neural Networks

Download

Quality Enhancement of Computed Tomography Images of Porous Media Using Convolutional Neural Networks.pdf

Date

2022-2-11

Author

Yıldırım, Ertuğrul Umut

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

325
views

131
downloads

Computed tomography has been widely used in clinical and industrial applications as a non-destructive visualization technology. The quality of computed tomography scans has a strong effect on the accuracy of the estimated physical properties of the investigated sample. X-ray exposure time is a crucial factor for scan quality. Ideally, long exposure time scans, yielding large signal-to-noise ratios, are available if physical properties are to be delineated. However, especially in micro-computed tomography applications, long exposure times constitute a problem for monitoring some physical processes that are happening quickly. To alleviate this problem, this thesis proposes a convolutional neural network approach for scan quality enhancement allowing for a reduction in X-ray exposure time while improving signal-to-noise ratio of the scanned image simultaneously. Moreover, the impact of using different loss functions, namely the mean squared error and the structural similarity index measure, on the performance of the network is analyzed. Both the visual and quantitative assessments show that the trained network greatly improves the quality of low-dose scans.

Subject Keywords

convolutional neural networks, computed tomography, denoising

URI

https://hdl.handle.net/11511/96710

Collections

Graduate School of Applied Mathematics, Thesis

Suggestions

OpenMETU
Core

ELECTRICAL-IMPEDANCE TOMOGRAPHY USING INDUCED CURRENTS Gençer, Nevzat Güneri; Kuzuoğlu, Mustafa (1994-06-01) The mathematical basis of a new imaging modality, Induced Current Electrical Impedance Tomography (EIT), is investigated. The ultimate aim of this technique is the reconstruction of conductivity distribution of the human body, from voltage measurements made between electrodes placed on the surface, when currents are induced inside the body by applied time varying magnetic fields. In this study the two-dimensional problem is analyzed. A specific 9-coil system for generating nine different exciting magnetic f...
Parameter estimation for instantaneous spectral imaging Öktem, Sevinç Figen; Davila, Joseph M (2014-05-04) Spectral imaging is a fundamental diagnostic technique in physical sciences with widespread application. Conventionally, spectral imaging techniques rely on a scanning process, which renders them unsuitable for dynamic scenes. Here we study the problem of estimating the physical parameters of interest from the measurements of a non-scanning spectral imager based on a parametric model. This inverse problem, which can be viewed as a multi-frame deblurring problem, is formulated as a maximum a posteriori (MAP)...
Neural networks with piecewise constant argument and impact activation Yılmaz, Enes; Akhmet, Marat; Department of Scientific Computing (2011) This dissertation addresses the new models in mathematical neuroscience: artificial neural networks, which have many similarities with the structure of human brain and the functions of cells by electronic circuits. The networks have been investigated due to their extensive applications in classification of patterns, associative memories, image processing, artificial intelligence, signal processing and optimization problems. These applications depend crucially on the dynamical behaviors of the networks. In t...
Cramer Rao bounds and instrument optimization for slitless spectroscopy Öktem, Sevinç Figen; Davila, Joseph (2013-05-26) Spectroscopy is a fundamental diagnostic technique in physical sciences with widespread application. Multi-order slitless imaging spectroscopy has been recently proposed to overcome the limitations of traditional spectrographs, in particular their small instantaneous field of view. Since an inversion is required to infer the physical parameters of interest from slitless spectroscopic measurements, a rigorous theory is essential for quantitative characterization of their performance. In this paper we develop...
Data acquisition system for MAET with magnetic field measurements Kaboutari, Keivan; Tetik, Ahmet Onder; Ghalichi, Elyar; Gozu, Mehmet Soner; Zengin, Reyhan; Gençer, Nevzat Güneri (2019-06-01) Magneto-acousto-electrical tomography (MAET) is an imaging modality to image the electrical conductivity of biological tissues. It is based on electrical current induction by using ultrasound under a static magnetic field. The aim of this study is to develop a data acquisition system for MAET based on magnetic field measurements. The static magnetic field is generated by six permanent neodymium magnets. A 16-element linear phased array (LPA) transducer is utilized to generate acoustic pressure waves inside ...

Citation Formats

E. U. Yıldırım, “Quality Enhancement of Computed Tomography Images of Porous Media Using Convolutional Neural Networks,” M.S. - Master of Science, Middle East Technical University, 2022.