Geodesic Information Flows: Spatially-Variant Graphs and Their Application to Segmentation and Fusion

M. Jorge Cardoso; Marc Modat; Robin Wolz; Andrew Melbourne; David Cash; Daniel Rueckert; Sebastien Ourselin

doi:10.1109/TMI.2015.2418298

Geodesic Information Flows: Spatially-Variant Graphs and Their Application to Segmentation and Fusion

M. Jorge Cardoso^*, Marc Modat, Robin Wolz, Andrew Melbourne, David Cash, Daniel Rueckert, Sebastien Ourselin

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

259 Citations (Scopus)

171 Downloads (Pure)

Abstract

Clinical annotations, such as voxel-wise binary or probabilistic tissue segmentations, structural parcellations, pathological regions-of-interest and anatomical landmarks are key to many clinical studies. However, due to the time consuming nature of manually generating these annotations, they tend to be scarce and limited to small subsets of data. This work explores a novel framework to propagate voxel-wise annotations between morphologically dissimilar images by diffusing and mapping the available examples through intermediate steps. A spatially-variant graph structure connecting morphologically similar subjects is introduced over a database of images, enabling the gradual diffusion of information to all the subjects, even in the presence of large-scale morphological variability. We illustrate the utility of the proposed framework on two example applications: brain parcellation using categorical labels and tissue segmentation using probabilistic features. The application of the proposed method to categorical label fusion showed highly statistically significant improvements when compared to state-of-the-art methodologies. Significant improvements were also observed when applying the proposed framework to probabilistic tissue segmentation of both synthetic and real data, mainly in the presence of large morphological variability.

Original language	English
Article number	7086081
Pages (from-to)	1976-1988
Number of pages	13
Journal	IEEE Transactions on Medical Imaging
Volume	34
Issue number	9
Early online date	14 Apr 2015
DOIs	https://doi.org/10.1109/TMI.2015.2418298
Publication status	Published - 1 Sept 2015

Keywords

Information propagation
label fusion
parcelation
tissue segmentation

Access to Document

10.1109/TMI.2015.2418298Licence: CC BY

Geodesic Information Flows_CARDOSO_Publishedonline14April2015_GOLD VoR (CC BY)Final published version, 1.73 MBLicence: CC BY

Cite this

@article{5a76f51ad62a42fe9c42973662bc87f9,

title = "Geodesic Information Flows: Spatially-Variant Graphs and Their Application to Segmentation and Fusion",

abstract = "Clinical annotations, such as voxel-wise binary or probabilistic tissue segmentations, structural parcellations, pathological regions-of-interest and anatomical landmarks are key to many clinical studies. However, due to the time consuming nature of manually generating these annotations, they tend to be scarce and limited to small subsets of data. This work explores a novel framework to propagate voxel-wise annotations between morphologically dissimilar images by diffusing and mapping the available examples through intermediate steps. A spatially-variant graph structure connecting morphologically similar subjects is introduced over a database of images, enabling the gradual diffusion of information to all the subjects, even in the presence of large-scale morphological variability. We illustrate the utility of the proposed framework on two example applications: brain parcellation using categorical labels and tissue segmentation using probabilistic features. The application of the proposed method to categorical label fusion showed highly statistically significant improvements when compared to state-of-the-art methodologies. Significant improvements were also observed when applying the proposed framework to probabilistic tissue segmentation of both synthetic and real data, mainly in the presence of large morphological variability.",

keywords = "Information propagation, label fusion, parcelation, tissue segmentation",

author = "Cardoso, {M. Jorge} and Marc Modat and Robin Wolz and Andrew Melbourne and David Cash and Daniel Rueckert and Sebastien Ourselin",

year = "2015",

month = sep,

day = "1",

doi = "10.1109/TMI.2015.2418298",

language = "English",

volume = "34",

pages = "1976--1988",

journal = "IEEE Transactions on Medical Imaging",

issn = "0278-0062",

publisher = "I E E E",

number = "9",

}

TY - JOUR

T1 - Geodesic Information Flows

T2 - Spatially-Variant Graphs and Their Application to Segmentation and Fusion

AU - Cardoso, M. Jorge

AU - Modat, Marc

AU - Wolz, Robin

AU - Melbourne, Andrew

AU - Cash, David

AU - Rueckert, Daniel

AU - Ourselin, Sebastien

PY - 2015/9/1

Y1 - 2015/9/1

N2 - Clinical annotations, such as voxel-wise binary or probabilistic tissue segmentations, structural parcellations, pathological regions-of-interest and anatomical landmarks are key to many clinical studies. However, due to the time consuming nature of manually generating these annotations, they tend to be scarce and limited to small subsets of data. This work explores a novel framework to propagate voxel-wise annotations between morphologically dissimilar images by diffusing and mapping the available examples through intermediate steps. A spatially-variant graph structure connecting morphologically similar subjects is introduced over a database of images, enabling the gradual diffusion of information to all the subjects, even in the presence of large-scale morphological variability. We illustrate the utility of the proposed framework on two example applications: brain parcellation using categorical labels and tissue segmentation using probabilistic features. The application of the proposed method to categorical label fusion showed highly statistically significant improvements when compared to state-of-the-art methodologies. Significant improvements were also observed when applying the proposed framework to probabilistic tissue segmentation of both synthetic and real data, mainly in the presence of large morphological variability.

AB - Clinical annotations, such as voxel-wise binary or probabilistic tissue segmentations, structural parcellations, pathological regions-of-interest and anatomical landmarks are key to many clinical studies. However, due to the time consuming nature of manually generating these annotations, they tend to be scarce and limited to small subsets of data. This work explores a novel framework to propagate voxel-wise annotations between morphologically dissimilar images by diffusing and mapping the available examples through intermediate steps. A spatially-variant graph structure connecting morphologically similar subjects is introduced over a database of images, enabling the gradual diffusion of information to all the subjects, even in the presence of large-scale morphological variability. We illustrate the utility of the proposed framework on two example applications: brain parcellation using categorical labels and tissue segmentation using probabilistic features. The application of the proposed method to categorical label fusion showed highly statistically significant improvements when compared to state-of-the-art methodologies. Significant improvements were also observed when applying the proposed framework to probabilistic tissue segmentation of both synthetic and real data, mainly in the presence of large morphological variability.

KW - Information propagation

KW - label fusion

KW - parcelation

KW - tissue segmentation

UR - http://www.scopus.com/inward/record.url?scp=84940912367&partnerID=8YFLogxK

U2 - 10.1109/TMI.2015.2418298

DO - 10.1109/TMI.2015.2418298

M3 - Article

C2 - 25879909

AN - SCOPUS:84940912367

SN - 0278-0062

VL - 34

SP - 1976

EP - 1988

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

IS - 9

M1 - 7086081

ER -

Geodesic Information Flows: Spatially-Variant Graphs and Their Application to Segmentation and Fusion

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this