2017
DOI: 10.1002/mrm.26841
|View full text |Cite
|
Sign up to set email alerts
|

Abstract: Purpose: To describe and evaluate a new fully automated musculoskeletal tissue segmentation method using deep convolutional neural network (CNN) and three-dimensional (3D) simplex deformable modeling to improve the accuracy and efficiency of cartilage and bone segmentation within the knee joint. Methods: A fully automated segmentation pipeline was built by combining a semantic segmentation CNN and 3D simplex deformable modeling. A CNN technique called SegNet was applied as the core of the segmentation method… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

3
224
0
2

Year Published

2018
2018
2022
2022

Publication Types

Select...
9

Relationship

2
7

Authors

Journals

citations
Cited by 254 publications
(229 citation statements)
references
References 56 publications
3
224
0
2
Order By: Relevance
“…The neural network model chosen for this problem is based on the U-Net architecture, which has previously shown promising results in the tasks of segmentation, particularly for medical images (15,2225), and has fewer trainable parameters than the other popular segmentation architecture, SegNet (26). The U-Net architecture can be viewed in Figure E1 (online).…”
Section: Methodsmentioning
confidence: 99%
“…The neural network model chosen for this problem is based on the U-Net architecture, which has previously shown promising results in the tasks of segmentation, particularly for medical images (15,2225), and has fewer trainable parameters than the other popular segmentation architecture, SegNet (26). The U-Net architecture can be viewed in Figure E1 (online).…”
Section: Methodsmentioning
confidence: 99%
“…The DL framework for lung segmentation (Fig. a) used a 2D convolutional encoder‐decoder (CED) architecture, which has been successfully applied for cartilage and brain tissue segmentation . The encoder network uses the same 13 Visual Geometry Group 16 convolutional layers and the decoder uses a mirrored structure of the encoder network with max‐pooling replaced by an upsampling process.…”
Section: Methodsmentioning
confidence: 99%
“…These types of networks, which have proven to be successful in related problem contexts (Ronneberger et al, 2015;Liu et al, 2018), transform the input image into a latent space, where, in our case, abstract representations of the fibril images are obtained at lower resolution, before reprojecting them onto the original resolution of the image. For this type of problem, fully CNNs are well established, especially when dealing with complex data as given by the varying appearance of crossovers in microscopic image data.…”
Section: Description Of the Cnn Architecturementioning
confidence: 99%
“…In particular, in recent years, encoder-decoder architectures (Ronneberger et al, 2015;Liu et al, 2018) were established for this kind of problems. In particular, in recent years, encoder-decoder architectures (Ronneberger et al, 2015;Liu et al, 2018) were established for this kind of problems.…”
Section: Introductionmentioning
confidence: 99%