Medical Image Denoising with Deep Convolutional Neural Networks

Medical Image Denoising with Deep Convolutional Neural Networks

Presenter(s)

Zahangir Alom

Description

In the last few years, Deep Leaning (DL) approaches are applied in different modalities of Bio-Medical imaging application including classification, segmentation, and detection tasks. In addition, DL based generative methods are also used for image denoising and restoration tasks. In particular, the generative models have applied for enhancement and restoration of Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) images and achieved state-of-the-art performance for noise cancelation and restoration. In this work, we apply different generative model including Generative Adversarial Network (GAN), and denoising convolutional auto-encoder for bio-medical image enhancement problem. The experiments are conducted on different publicly available datasets for MRI and CT images. The experimental result shows promising outputs which can be applied for different applications in the modalities of MRI and CT.

Publication Date

4-24-2019

Project Designation

Independent Research

Primary Advisor

Tarek M. Taha, Vijayan K. Asari

Primary Advisor's Department

Electrical and Computer Engineering

Keywords

Stander Symposium project

Medical Image Denoising with Deep Convolutional Neural Networks

Share

COinS