USDL: Inexpensive Medical Imaging Using Deep Learning Techniques and Ultrasound Technology.
Autoencoders
Deep Learning
Denoising
In Vivo Ultrasounds
MSE
PSNR
SSIM
Speckle Noise
Ultrasound imaging
Journal
2020 Design of Medical Devices Conference (DMD 2020). Design of Medical Devices Conferences (2020 : Minneapolis, Minn.)
Titre abrégé: Des Med Devices Conf (2020)
Pays: United States
ID NLM: 9918350484006676
Informations de publication
Date de publication:
Apr 2020
Apr 2020
Historique:
entrez:
7
3
2022
pubmed:
1
4
2020
medline:
1
4
2020
Statut:
ppublish
Résumé
In this study, we present USDL, a novel model that employs deep learning algorithms in order to reconstruct and enhance corrupted ultrasound images. We utilize an unsupervised neural network called an autoencoder which works by compressing its input into a latent-space representation and then reconstructing the output from this representation. We trained our model on a dataset that compromises of 15,700
Identifiants
pubmed: 35253013
doi: 10.1115/dmd2020-9109
pmc: PMC8895229
mid: NIHMS1779209
pii:
doi:
Types de publication
Journal Article
Langues
eng
Subventions
Organisme : NCATS NIH HHS
ID : KL2 TR003099
Pays : United States
Références
Neuroimage. 2016 Jan 1;124(Pt A):752-761
pubmed: 26416649
Med Image Anal. 2017 Dec;42:60-88
pubmed: 28778026
IEEE Trans Med Imaging. 2018 Sep;37(9):2010-2021
pubmed: 29994441