Combining outlier detection with random walker for automatic brain tumor segmentation

Vasileios G. Kanas, Evangelia I. Zacharaki, Evangelos Dermatas, Anastasios Bezerianos, Kyriakos Sgarbas, Christos Davatzikos

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

Abstract

The diagnosis of brain neoplasms has been facilitated by the emerging of high-quality imaging techniques, such as Magnetic Resonance Imaging (MRI), while the combination of several sequences from conventional and advanced protocols has increased the diagnostic information. Treatment planning and therapy follow-up require the detection of neoplastic and edematous tissue boundaries, a very time consuming task when manually performed by medical experts based on the 3D MRI data. Automating the detection process is challenging, due to the high diversity in appearance of neoplastic tissue among different patients and, in many cases, similarity between neoplastic and normal tissue. In this paper, we propose an automatic brain tumor segmentation method based on a multilabel multiparametric random walks approach initialized by an outlier detection scheme. Segmentation assessment is performed by measuring spatial overlap between automatic segmentation and manual segmentation performed by medical experts. Good agreement is observed in most of the 26 cases for both neoplastic and edematous tissue. The highest achieved overlapping values were 0.74 and 0.79 for neoplastic and edematous tissue, respectively.

Original languageEnglish (US)
Title of host publicationIFIP Advances in Information and Communication Technology
Pages26-35
Number of pages10
Volume382 AICT
EditionPART 2
DOIs
StatePublished - 2012
Externally publishedYes
Event8th International Workshop on Artificial Intelligence Applications and Innovations, AIAI 2012: AIAB, AIeIA, CISE, COPA, IIVC, ISQL, MHDW, and WADTMB - Halkidiki, Greece
Duration: Sep 27 2012Sep 30 2012

Publication series

NameIFIP Advances in Information and Communication Technology
NumberPART 2
Volume382 AICT
ISSN (Print)18684238

Other

Other8th International Workshop on Artificial Intelligence Applications and Innovations, AIAI 2012: AIAB, AIeIA, CISE, COPA, IIVC, ISQL, MHDW, and WADTMB
CountryGreece
CityHalkidiki
Period9/27/129/30/12

Fingerprint

Tumor
Outlier detection
Segmentation
Magnetic resonance imaging
Diagnostics
Therapy
Planning
Random walk
Overlapping
Imaging

Keywords

  • brain neoplasms
  • k- means
  • magnetic resonance imaging
  • outlier detection
  • random walks
  • segmentation

ASJC Scopus subject areas

  • Information Systems and Management

Cite this

Kanas, V. G., Zacharaki, E. I., Dermatas, E., Bezerianos, A., Sgarbas, K., & Davatzikos, C. (2012). Combining outlier detection with random walker for automatic brain tumor segmentation. In IFIP Advances in Information and Communication Technology (PART 2 ed., Vol. 382 AICT, pp. 26-35). (IFIP Advances in Information and Communication Technology; Vol. 382 AICT, No. PART 2). https://doi.org/10.1007/978-3-642-33412-2_3

Combining outlier detection with random walker for automatic brain tumor segmentation. / Kanas, Vasileios G.; Zacharaki, Evangelia I.; Dermatas, Evangelos; Bezerianos, Anastasios; Sgarbas, Kyriakos; Davatzikos, Christos.

IFIP Advances in Information and Communication Technology. Vol. 382 AICT PART 2. ed. 2012. p. 26-35 (IFIP Advances in Information and Communication Technology; Vol. 382 AICT, No. PART 2).

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

Kanas, VG, Zacharaki, EI, Dermatas, E, Bezerianos, A, Sgarbas, K & Davatzikos, C 2012, Combining outlier detection with random walker for automatic brain tumor segmentation. in IFIP Advances in Information and Communication Technology. PART 2 edn, vol. 382 AICT, IFIP Advances in Information and Communication Technology, no. PART 2, vol. 382 AICT, pp. 26-35, 8th International Workshop on Artificial Intelligence Applications and Innovations, AIAI 2012: AIAB, AIeIA, CISE, COPA, IIVC, ISQL, MHDW, and WADTMB, Halkidiki, Greece, 9/27/12. https://doi.org/10.1007/978-3-642-33412-2_3
Kanas VG, Zacharaki EI, Dermatas E, Bezerianos A, Sgarbas K, Davatzikos C. Combining outlier detection with random walker for automatic brain tumor segmentation. In IFIP Advances in Information and Communication Technology. PART 2 ed. Vol. 382 AICT. 2012. p. 26-35. (IFIP Advances in Information and Communication Technology; PART 2). https://doi.org/10.1007/978-3-642-33412-2_3
Kanas, Vasileios G. ; Zacharaki, Evangelia I. ; Dermatas, Evangelos ; Bezerianos, Anastasios ; Sgarbas, Kyriakos ; Davatzikos, Christos. / Combining outlier detection with random walker for automatic brain tumor segmentation. IFIP Advances in Information and Communication Technology. Vol. 382 AICT PART 2. ed. 2012. pp. 26-35 (IFIP Advances in Information and Communication Technology; PART 2).
@inproceedings{131ad829897144d28e328f2a6658ffa7,
title = "Combining outlier detection with random walker for automatic brain tumor segmentation",
abstract = "The diagnosis of brain neoplasms has been facilitated by the emerging of high-quality imaging techniques, such as Magnetic Resonance Imaging (MRI), while the combination of several sequences from conventional and advanced protocols has increased the diagnostic information. Treatment planning and therapy follow-up require the detection of neoplastic and edematous tissue boundaries, a very time consuming task when manually performed by medical experts based on the 3D MRI data. Automating the detection process is challenging, due to the high diversity in appearance of neoplastic tissue among different patients and, in many cases, similarity between neoplastic and normal tissue. In this paper, we propose an automatic brain tumor segmentation method based on a multilabel multiparametric random walks approach initialized by an outlier detection scheme. Segmentation assessment is performed by measuring spatial overlap between automatic segmentation and manual segmentation performed by medical experts. Good agreement is observed in most of the 26 cases for both neoplastic and edematous tissue. The highest achieved overlapping values were 0.74 and 0.79 for neoplastic and edematous tissue, respectively.",
keywords = "brain neoplasms, k- means, magnetic resonance imaging, outlier detection, random walks, segmentation",
author = "Kanas, {Vasileios G.} and Zacharaki, {Evangelia I.} and Evangelos Dermatas and Anastasios Bezerianos and Kyriakos Sgarbas and Christos Davatzikos",
year = "2012",
doi = "10.1007/978-3-642-33412-2_3",
language = "English (US)",
isbn = "9783642334115",
volume = "382 AICT",
series = "IFIP Advances in Information and Communication Technology",
number = "PART 2",
pages = "26--35",
booktitle = "IFIP Advances in Information and Communication Technology",
edition = "PART 2",

}

TY - GEN

T1 - Combining outlier detection with random walker for automatic brain tumor segmentation

AU - Kanas, Vasileios G.

AU - Zacharaki, Evangelia I.

AU - Dermatas, Evangelos

AU - Bezerianos, Anastasios

AU - Sgarbas, Kyriakos

AU - Davatzikos, Christos

PY - 2012

Y1 - 2012

N2 - The diagnosis of brain neoplasms has been facilitated by the emerging of high-quality imaging techniques, such as Magnetic Resonance Imaging (MRI), while the combination of several sequences from conventional and advanced protocols has increased the diagnostic information. Treatment planning and therapy follow-up require the detection of neoplastic and edematous tissue boundaries, a very time consuming task when manually performed by medical experts based on the 3D MRI data. Automating the detection process is challenging, due to the high diversity in appearance of neoplastic tissue among different patients and, in many cases, similarity between neoplastic and normal tissue. In this paper, we propose an automatic brain tumor segmentation method based on a multilabel multiparametric random walks approach initialized by an outlier detection scheme. Segmentation assessment is performed by measuring spatial overlap between automatic segmentation and manual segmentation performed by medical experts. Good agreement is observed in most of the 26 cases for both neoplastic and edematous tissue. The highest achieved overlapping values were 0.74 and 0.79 for neoplastic and edematous tissue, respectively.

AB - The diagnosis of brain neoplasms has been facilitated by the emerging of high-quality imaging techniques, such as Magnetic Resonance Imaging (MRI), while the combination of several sequences from conventional and advanced protocols has increased the diagnostic information. Treatment planning and therapy follow-up require the detection of neoplastic and edematous tissue boundaries, a very time consuming task when manually performed by medical experts based on the 3D MRI data. Automating the detection process is challenging, due to the high diversity in appearance of neoplastic tissue among different patients and, in many cases, similarity between neoplastic and normal tissue. In this paper, we propose an automatic brain tumor segmentation method based on a multilabel multiparametric random walks approach initialized by an outlier detection scheme. Segmentation assessment is performed by measuring spatial overlap between automatic segmentation and manual segmentation performed by medical experts. Good agreement is observed in most of the 26 cases for both neoplastic and edematous tissue. The highest achieved overlapping values were 0.74 and 0.79 for neoplastic and edematous tissue, respectively.

KW - brain neoplasms

KW - k- means

KW - magnetic resonance imaging

KW - outlier detection

KW - random walks

KW - segmentation

UR - http://www.scopus.com/inward/record.url?scp=84870904188&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84870904188&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-33412-2_3

DO - 10.1007/978-3-642-33412-2_3

M3 - Conference contribution

AN - SCOPUS:84870904188

SN - 9783642334115

VL - 382 AICT

T3 - IFIP Advances in Information and Communication Technology

SP - 26

EP - 35

BT - IFIP Advances in Information and Communication Technology

ER -