Files in this item



application/pdfCOLE-THESIS-2020.pdf (4MB)Restricted to U of Illinois
(no description provided)PDF


Title:Joint super resolution and denoising: learning to recover sharp features in radiology images
Author(s):Cole, Patrick Alexander
Advisor(s):Koyejo, Oluwasanmi
Department / Program:Computer Science
Discipline:Computer Science
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Machine Learning
Artificial Intelligence
Computer Vision
X-ray Radiograph
Computed Tomography
Super Resolution
Image Processing
Abstract:Radiology exams require exposing a patient to a variable dosage of radiation. The amount of radiation used during the exam directly corresponds to the level of noise in the resulting image. While large amounts of radiation can be dangerous for certain patients, radiologists need an uncorrupted image to make a diagnosis. In our work, we detail methods for simulating low-dose noise for two popular radiology exams: x-ray radiograph and computed tomography. We propose a methodology to recover the uncorrupted exam results given a noisy, or low-dose, sample. Using a two-part criterion that consists of a pixel-wise loss and an adversarial loss, we are able to recover the structure and fine detail of the normal-dose sample.
Issue Date:2020-07-22
Rights Information:Copyright 2020 Patrick Cole
Date Available in IDEALS:2020-10-07
Date Deposited:2020-08

This item appears in the following Collection(s)

Item Statistics