Medical Image Denoising

2021 Research Internship Project


Faculty Name

Javad Alirezaie

Project Title

Medical Image Denoising

Project Description

One of the most useful tools in clinical application for diagnostic radiology is Computed Tomography (CT). While CT scans are a terrific diagnostic resource in health care, there are concerns about the risk of radiation as CT scans can deliver up to 200 times the radiation that single X-Ray can. Increasing exposure to radiation via this imaging modality in the population is considered a public health challenge. Image post processing techniques to enhance the quality of noisy images taken at low doses can help minimize radiation exposure with little sacrifice in image quality. In my CVIP lab, we are proposing several denoising methods utilizing deep learning and sparse representations to address the challenge of low-dose CT image denoising. Deep learning is a branch of machine learning, and the goal is to learn the representation of the data. During the training process, the algorithm observes and processes many samples and finds the distribution of the target with respect to the input data. It is especially beneficial in identifying the patterns of unstructured data, similar to the problem of this study. Stacking more layers in neural networks and developing techniques to train and improve the performance has led to deep learning. Our research aims to develop and test different techniques, as mentioned above, to address the denoising of low-dose CT images.

Student Responsibility

The student’s work will be focused on the utilization of various statistical and machine learning algorithms to achieve an accurate representation of the given dataset. With this, they will work towards optimizing the deep learning approach taken to achieve accurate and meaningful results. They will be responsible for implementing the developed techniques in the lab for denoising the dataset of CT images. They will also assist in the creation of reports, graphs, and statistics to contribute to graduate student work with the development of manuscripts for publication.

Specific Requirements

1) Knowledge of Matlab and/or Python. 2) Knowledge of Signals and Systems I and II 3) Basic knowledge of image processing will be helpful but not strictly required.

Application Procedure

To apply for this project you will need to login to the departmental web portal and select Research Internship from the sidebar menu.

Portal Login

Javad Alirezaie : Medical Image Denoising | Thursday April 1st 2021 02:02 PM