A novel hierarchical level set with AR-boost for white matter lesion segmentation in diabetes

Baidya N. Saha, Sriraam Natarajan, Gopi Kota, Christopher T. Whitlow, Donald Bowden, Jasmin Divers, Barry I. Freedman, Joseph A Maldjian

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

1 Citation (Scopus)

Abstract

Hierarchical as well as coupled level sets are widely used for multilevel image segmentation. However, these tools are successful if the number of levels of an image are known and a careful choice of initialization is performed. We intend a novel hierarchical level set (HLS) followed by an Adaptive Regularized Boosting (AR-Boost) for automatic White Matter Lesion (WML) segmentation from Magnetic Resonance Images. HLS does not need to know the number of levels in an image and HLS is computationally less expensive and more initialization independent than coupled level-setssince HLS doesn't generate redundant regions. We employan energy functional that minimizes the negative logarithm of variances between the two partitions created by the level set function. HLS uses a level set to partition the image into a number of segments, then applies the level set on all the segments separately to create more segments and the process continues iteratively until all the segments become a nearly homogeneous region (low intensity variance). Then AR-Boost classifies the segments into WML and non-WML classes. The proposed loss function for AR-boost enforces more weight on misclassified samples at each iteration than Adaboost to classify correctly in the next iteration and consequently leads to early convergence. Unlike Adaboost, the user can select optimal weights through cross-validation. Experimental results demonstrate that the proposed method outperforms state-of-the-art automated white matter lesion segmentation techniques.

Original languageEnglish (US)
Title of host publicationProceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012
Pages90-95
Number of pages6
Volume1
DOIs
StatePublished - Dec 1 2012
Event11th IEEE International Conference on Machine Learning and Applications, ICMLA 2012 - Boca Raton, FL, United States
Duration: Dec 12 2012Dec 15 2012

Other

Other11th IEEE International Conference on Machine Learning and Applications, ICMLA 2012
CountryUnited States
CityBoca Raton, FL
Period12/12/1212/15/12

Fingerprint

Adaptive boosting
Medical problems
chronic illness
Magnetic resonance
Image segmentation
energy
segmentation

Keywords

  • Boosting
  • Diabetes
  • Level Set
  • White Matter Lesion Segmentation

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Education

Cite this

Saha, B. N., Natarajan, S., Kota, G., Whitlow, C. T., Bowden, D., Divers, J., ... Maldjian, J. A. (2012). A novel hierarchical level set with AR-boost for white matter lesion segmentation in diabetes. In Proceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012 (Vol. 1, pp. 90-95). [6406594] https://doi.org/10.1109/ICMLA.2012.24

A novel hierarchical level set with AR-boost for white matter lesion segmentation in diabetes. / Saha, Baidya N.; Natarajan, Sriraam; Kota, Gopi; Whitlow, Christopher T.; Bowden, Donald; Divers, Jasmin; Freedman, Barry I.; Maldjian, Joseph A.

Proceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012. Vol. 1 2012. p. 90-95 6406594.

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

Saha, BN, Natarajan, S, Kota, G, Whitlow, CT, Bowden, D, Divers, J, Freedman, BI & Maldjian, JA 2012, A novel hierarchical level set with AR-boost for white matter lesion segmentation in diabetes. in Proceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012. vol. 1, 6406594, pp. 90-95, 11th IEEE International Conference on Machine Learning and Applications, ICMLA 2012, Boca Raton, FL, United States, 12/12/12. https://doi.org/10.1109/ICMLA.2012.24
Saha BN, Natarajan S, Kota G, Whitlow CT, Bowden D, Divers J et al. A novel hierarchical level set with AR-boost for white matter lesion segmentation in diabetes. In Proceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012. Vol. 1. 2012. p. 90-95. 6406594 https://doi.org/10.1109/ICMLA.2012.24
Saha, Baidya N. ; Natarajan, Sriraam ; Kota, Gopi ; Whitlow, Christopher T. ; Bowden, Donald ; Divers, Jasmin ; Freedman, Barry I. ; Maldjian, Joseph A. / A novel hierarchical level set with AR-boost for white matter lesion segmentation in diabetes. Proceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012. Vol. 1 2012. pp. 90-95
@inproceedings{c36fb6484ab04450b89fbaf7900ad113,
title = "A novel hierarchical level set with AR-boost for white matter lesion segmentation in diabetes",
abstract = "Hierarchical as well as coupled level sets are widely used for multilevel image segmentation. However, these tools are successful if the number of levels of an image are known and a careful choice of initialization is performed. We intend a novel hierarchical level set (HLS) followed by an Adaptive Regularized Boosting (AR-Boost) for automatic White Matter Lesion (WML) segmentation from Magnetic Resonance Images. HLS does not need to know the number of levels in an image and HLS is computationally less expensive and more initialization independent than coupled level-setssince HLS doesn't generate redundant regions. We employan energy functional that minimizes the negative logarithm of variances between the two partitions created by the level set function. HLS uses a level set to partition the image into a number of segments, then applies the level set on all the segments separately to create more segments and the process continues iteratively until all the segments become a nearly homogeneous region (low intensity variance). Then AR-Boost classifies the segments into WML and non-WML classes. The proposed loss function for AR-boost enforces more weight on misclassified samples at each iteration than Adaboost to classify correctly in the next iteration and consequently leads to early convergence. Unlike Adaboost, the user can select optimal weights through cross-validation. Experimental results demonstrate that the proposed method outperforms state-of-the-art automated white matter lesion segmentation techniques.",
keywords = "Boosting, Diabetes, Level Set, White Matter Lesion Segmentation",
author = "Saha, {Baidya N.} and Sriraam Natarajan and Gopi Kota and Whitlow, {Christopher T.} and Donald Bowden and Jasmin Divers and Freedman, {Barry I.} and Maldjian, {Joseph A}",
year = "2012",
month = "12",
day = "1",
doi = "10.1109/ICMLA.2012.24",
language = "English (US)",
isbn = "9780769549132",
volume = "1",
pages = "90--95",
booktitle = "Proceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012",

}

TY - GEN

T1 - A novel hierarchical level set with AR-boost for white matter lesion segmentation in diabetes

AU - Saha, Baidya N.

AU - Natarajan, Sriraam

AU - Kota, Gopi

AU - Whitlow, Christopher T.

AU - Bowden, Donald

AU - Divers, Jasmin

AU - Freedman, Barry I.

AU - Maldjian, Joseph A

PY - 2012/12/1

Y1 - 2012/12/1

N2 - Hierarchical as well as coupled level sets are widely used for multilevel image segmentation. However, these tools are successful if the number of levels of an image are known and a careful choice of initialization is performed. We intend a novel hierarchical level set (HLS) followed by an Adaptive Regularized Boosting (AR-Boost) for automatic White Matter Lesion (WML) segmentation from Magnetic Resonance Images. HLS does not need to know the number of levels in an image and HLS is computationally less expensive and more initialization independent than coupled level-setssince HLS doesn't generate redundant regions. We employan energy functional that minimizes the negative logarithm of variances between the two partitions created by the level set function. HLS uses a level set to partition the image into a number of segments, then applies the level set on all the segments separately to create more segments and the process continues iteratively until all the segments become a nearly homogeneous region (low intensity variance). Then AR-Boost classifies the segments into WML and non-WML classes. The proposed loss function for AR-boost enforces more weight on misclassified samples at each iteration than Adaboost to classify correctly in the next iteration and consequently leads to early convergence. Unlike Adaboost, the user can select optimal weights through cross-validation. Experimental results demonstrate that the proposed method outperforms state-of-the-art automated white matter lesion segmentation techniques.

AB - Hierarchical as well as coupled level sets are widely used for multilevel image segmentation. However, these tools are successful if the number of levels of an image are known and a careful choice of initialization is performed. We intend a novel hierarchical level set (HLS) followed by an Adaptive Regularized Boosting (AR-Boost) for automatic White Matter Lesion (WML) segmentation from Magnetic Resonance Images. HLS does not need to know the number of levels in an image and HLS is computationally less expensive and more initialization independent than coupled level-setssince HLS doesn't generate redundant regions. We employan energy functional that minimizes the negative logarithm of variances between the two partitions created by the level set function. HLS uses a level set to partition the image into a number of segments, then applies the level set on all the segments separately to create more segments and the process continues iteratively until all the segments become a nearly homogeneous region (low intensity variance). Then AR-Boost classifies the segments into WML and non-WML classes. The proposed loss function for AR-boost enforces more weight on misclassified samples at each iteration than Adaboost to classify correctly in the next iteration and consequently leads to early convergence. Unlike Adaboost, the user can select optimal weights through cross-validation. Experimental results demonstrate that the proposed method outperforms state-of-the-art automated white matter lesion segmentation techniques.

KW - Boosting

KW - Diabetes

KW - Level Set

KW - White Matter Lesion Segmentation

UR - http://www.scopus.com/inward/record.url?scp=84873575152&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84873575152&partnerID=8YFLogxK

U2 - 10.1109/ICMLA.2012.24

DO - 10.1109/ICMLA.2012.24

M3 - Conference contribution

AN - SCOPUS:84873575152

SN - 9780769549132

VL - 1

SP - 90

EP - 95

BT - Proceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012

ER -