Example based lesion segmentation

Snehashis Roy, Qing He, Aaron Carass, Amod Jog, Jennifer L. Cuzzocreo, Daniel S. Reich, Jerry Ladd Prince, Dzung Pham

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Automatic and accurate detection of white matter lesions is a significant step toward understanding the progression of many diseases, like Alzheimers disease or multiple sclerosis. Multi-modal MR images are often used to segment T2 white matter lesions that can represent regions of demyelination or ischemia. Some automated lesion segmentation methods describe the lesion intensities using generative models, and then classify the lesions with some combination of heuristics and cost minimization. In contrast, we propose a patch-based method, in which lesions are found using examples from an atlas containing multi-modal MR images and corresponding manual delineations of lesions. Patches from subject MR images are matched to patches from the atlas and lesion memberships are found based on patch similarity weights. We experiment on 43 subjects with MS, whose scans show various levels of lesion-load. We demonstrate significant improvement in Dice coefficient and total lesion volume compared to a state of the art model-based lesion segmentation method, indicating more accurate delineation of lesions.

Original languageEnglish (US)
Title of host publicationProgress in Biomedical Optics and Imaging - Proceedings of SPIE
PublisherSPIE
Volume9034
ISBN (Print)9780819498274
DOIs
StatePublished - 2014
EventMedical Imaging 2014: Image Processing - San Diego, CA, United States
Duration: Feb 16 2014Feb 18 2014

Other

OtherMedical Imaging 2014: Image Processing
CountryUnited States
CitySan Diego, CA
Period2/16/142/18/14

Fingerprint

lesions
Atlases
Alzheimer Disease
Demyelinating Diseases
Multiple Sclerosis
Disease Progression
Ischemia
Costs
Weights and Measures
Costs and Cost Analysis
delineation
Experiments
ischemia
White Matter
progressions
costs
optimization
Heuristics

Keywords

  • Lesion segmentation
  • Magnetic resonance imaging
  • MRI
  • MS
  • Patches

ASJC Scopus subject areas

  • Atomic and Molecular Physics, and Optics
  • Electronic, Optical and Magnetic Materials
  • Biomaterials
  • Radiology Nuclear Medicine and imaging

Cite this

Roy, S., He, Q., Carass, A., Jog, A., Cuzzocreo, J. L., Reich, D. S., ... Pham, D. (2014). Example based lesion segmentation. In Progress in Biomedical Optics and Imaging - Proceedings of SPIE (Vol. 9034). [90341Y] SPIE. https://doi.org/10.1117/12.2043917

Example based lesion segmentation. / Roy, Snehashis; He, Qing; Carass, Aaron; Jog, Amod; Cuzzocreo, Jennifer L.; Reich, Daniel S.; Prince, Jerry Ladd; Pham, Dzung.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE. Vol. 9034 SPIE, 2014. 90341Y.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Roy, S, He, Q, Carass, A, Jog, A, Cuzzocreo, JL, Reich, DS, Prince, JL & Pham, D 2014, Example based lesion segmentation. in Progress in Biomedical Optics and Imaging - Proceedings of SPIE. vol. 9034, 90341Y, SPIE, Medical Imaging 2014: Image Processing, San Diego, CA, United States, 2/16/14. https://doi.org/10.1117/12.2043917
Roy S, He Q, Carass A, Jog A, Cuzzocreo JL, Reich DS et al. Example based lesion segmentation. In Progress in Biomedical Optics and Imaging - Proceedings of SPIE. Vol. 9034. SPIE. 2014. 90341Y https://doi.org/10.1117/12.2043917
Roy, Snehashis ; He, Qing ; Carass, Aaron ; Jog, Amod ; Cuzzocreo, Jennifer L. ; Reich, Daniel S. ; Prince, Jerry Ladd ; Pham, Dzung. / Example based lesion segmentation. Progress in Biomedical Optics and Imaging - Proceedings of SPIE. Vol. 9034 SPIE, 2014.
@inproceedings{0c3d73ae693a451b8ead033b368239af,
title = "Example based lesion segmentation",
abstract = "Automatic and accurate detection of white matter lesions is a significant step toward understanding the progression of many diseases, like Alzheimers disease or multiple sclerosis. Multi-modal MR images are often used to segment T2 white matter lesions that can represent regions of demyelination or ischemia. Some automated lesion segmentation methods describe the lesion intensities using generative models, and then classify the lesions with some combination of heuristics and cost minimization. In contrast, we propose a patch-based method, in which lesions are found using examples from an atlas containing multi-modal MR images and corresponding manual delineations of lesions. Patches from subject MR images are matched to patches from the atlas and lesion memberships are found based on patch similarity weights. We experiment on 43 subjects with MS, whose scans show various levels of lesion-load. We demonstrate significant improvement in Dice coefficient and total lesion volume compared to a state of the art model-based lesion segmentation method, indicating more accurate delineation of lesions.",
keywords = "Lesion segmentation, Magnetic resonance imaging, MRI, MS, Patches",
author = "Snehashis Roy and Qing He and Aaron Carass and Amod Jog and Cuzzocreo, {Jennifer L.} and Reich, {Daniel S.} and Prince, {Jerry Ladd} and Dzung Pham",
year = "2014",
doi = "10.1117/12.2043917",
language = "English (US)",
isbn = "9780819498274",
volume = "9034",
booktitle = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",
publisher = "SPIE",

}

TY - GEN

T1 - Example based lesion segmentation

AU - Roy, Snehashis

AU - He, Qing

AU - Carass, Aaron

AU - Jog, Amod

AU - Cuzzocreo, Jennifer L.

AU - Reich, Daniel S.

AU - Prince, Jerry Ladd

AU - Pham, Dzung

PY - 2014

Y1 - 2014

N2 - Automatic and accurate detection of white matter lesions is a significant step toward understanding the progression of many diseases, like Alzheimers disease or multiple sclerosis. Multi-modal MR images are often used to segment T2 white matter lesions that can represent regions of demyelination or ischemia. Some automated lesion segmentation methods describe the lesion intensities using generative models, and then classify the lesions with some combination of heuristics and cost minimization. In contrast, we propose a patch-based method, in which lesions are found using examples from an atlas containing multi-modal MR images and corresponding manual delineations of lesions. Patches from subject MR images are matched to patches from the atlas and lesion memberships are found based on patch similarity weights. We experiment on 43 subjects with MS, whose scans show various levels of lesion-load. We demonstrate significant improvement in Dice coefficient and total lesion volume compared to a state of the art model-based lesion segmentation method, indicating more accurate delineation of lesions.

AB - Automatic and accurate detection of white matter lesions is a significant step toward understanding the progression of many diseases, like Alzheimers disease or multiple sclerosis. Multi-modal MR images are often used to segment T2 white matter lesions that can represent regions of demyelination or ischemia. Some automated lesion segmentation methods describe the lesion intensities using generative models, and then classify the lesions with some combination of heuristics and cost minimization. In contrast, we propose a patch-based method, in which lesions are found using examples from an atlas containing multi-modal MR images and corresponding manual delineations of lesions. Patches from subject MR images are matched to patches from the atlas and lesion memberships are found based on patch similarity weights. We experiment on 43 subjects with MS, whose scans show various levels of lesion-load. We demonstrate significant improvement in Dice coefficient and total lesion volume compared to a state of the art model-based lesion segmentation method, indicating more accurate delineation of lesions.

KW - Lesion segmentation

KW - Magnetic resonance imaging

KW - MRI

KW - MS

KW - Patches

UR - http://www.scopus.com/inward/record.url?scp=84902094483&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84902094483&partnerID=8YFLogxK

U2 - 10.1117/12.2043917

DO - 10.1117/12.2043917

M3 - Conference contribution

AN - SCOPUS:84902094483

SN - 9780819498274

VL - 9034

BT - Progress in Biomedical Optics and Imaging - Proceedings of SPIE

PB - SPIE

ER -