DUAL-ENERGY CBCT PRE-SPECTRAL-DECOMPOSITION FILTERING WITH WAVELETSHRINKAGE NETWORKS

Research output: Contribution to conferencePaperAcademic

Abstract

Convolutional Neural Networks (CNNs) are reshaping sig-nal processing and computer vision by providing data-drivensolutions for inverse problems such as noise reduction. How-ever, their relationship with established signal processingmethods is sometimes unclear and its development not fullyexploiting the existing knowledge. In this paper, rather thanimproving existing CNNs with wavelet transformations asexplored earlier, we improve the wavelet shrinkage approachto noise-reduction with a data-driven solution. The resultingCNN has clear encoding, decoding and processing paths.As application, we perform noise reduction in Dual-EnergyCone-Beam CT. The obtained results were compared to aUNet-like architecture, which reveal better noise-free imageswithout aliasing artifacts. This indicates that that our archi-tecture is able to preserve well the information contained inthe images because the architecture exploits explicitly theunderlying signal representation
Original languageEnglish
Pages1-6
Number of pages6
Publication statusPublished - 21 Sep 2020
EventInternational Workshop on Machine Learning for Signal Processing 2020 - Aalto University Campus, Espoo, Finland
Duration: 21 Sep 202024 Sep 2020
Conference number: 30th
https://ieeemlsp.cc/

Conference

ConferenceInternational Workshop on Machine Learning for Signal Processing 2020
Abbreviated titleMLSP 2020
CountryFinland
CityEspoo
Period21/09/2024/09/20
Internet address

Keywords

  • Convolutional neural networks, wavelets,noise-reduction, computer tomography

Fingerprint Dive into the research topics of 'DUAL-ENERGY CBCT PRE-SPECTRAL-DECOMPOSITION FILTERING WITH WAVELETSHRINKAGE NETWORKS'. Together they form a unique fingerprint.

Cite this