Skip to main content

Repeatability of Automated Image Segmentation with BraTumIA in Patients with Recurrent Glioblastoma.

Publication ,  Journal Article
Abu Khalaf, N; Desjardins, A; Vredenburgh, JJ; Barboriak, DP
Published in: AJNR Am J Neuroradiol
June 2021

BACKGROUND AND PURPOSE: Despite high interest in machine-learning algorithms for automated segmentation of MRIs of patients with brain tumors, there are few reports on the variability of segmentation results. The purpose of this study was to obtain benchmark measures of repeatability for a widely accessible software program, BraTumIA (Versions 1.2 and 2.0), which uses a machine-learning algorithm to segment tumor features on contrast-enhanced brain MR imaging. MATERIALS AND METHODS: Automatic segmentation of enhancing tumor, tumor edema, nonenhancing tumor, and necrosis was performed on repeat MR imaging scans obtained approximately 2 days apart in 20 patients with recurrent glioblastoma. Measures of repeatability and spatial overlap, including repeatability and Dice coefficients, are reported. RESULTS: Larger volumes of enhancing tumor were obtained on later compared with earlier scans (mean, 26.3 versus 24.2 mL for BraTumIA 1.2; P < .05; and 24.9 versus 22.9 mL for BraTumIA 2.0, P < .01). In terms of percentage change, repeatability coefficients ranged from 31% to 46% for enhancing tumor and edema components and from 87% to 116% for nonenhancing tumor and necrosis. Dice coefficients were highest (>0.7) for enhancing tumor and edema components, intermediate for necrosis, and lowest for nonenhancing tumor and did not differ between software versions. Enhancing tumor and tumor edema were smaller, and necrotic tumor larger using BraTumIA 2.0 rather than 1.2. CONCLUSIONS: Repeatability and overlap metrics varied by segmentation type, with better performance for segmentations of enhancing tumor and tumor edema compared with other components. Incomplete washout of gadolinium contrast agents could account for increasing enhancing tumor volumes on later scans.

Duke Scholars

Altmetric Attention Stats
Dimensions Citation Stats

Published In

AJNR Am J Neuroradiol

DOI

EISSN

1936-959X

Publication Date

June 2021

Volume

42

Issue

6

Start / End Page

1080 / 1086

Location

United States

Related Subject Headings

  • Tumor Burden
  • Nuclear Medicine & Medical Imaging
  • Magnetic Resonance Imaging
  • Image Processing, Computer-Assisted
  • Humans
  • Glioblastoma
  • Brain Neoplasms
  • Algorithms
  • 3406 Physical chemistry
  • 3209 Neurosciences
 

Citation

APA
Chicago
ICMJE
MLA
NLM
Abu Khalaf, N., Desjardins, A., Vredenburgh, J. J., & Barboriak, D. P. (2021). Repeatability of Automated Image Segmentation with BraTumIA in Patients with Recurrent Glioblastoma. AJNR Am J Neuroradiol, 42(6), 1080–1086. https://doi.org/10.3174/ajnr.A7071
Abu Khalaf, N., A. Desjardins, J. J. Vredenburgh, and D. P. Barboriak. “Repeatability of Automated Image Segmentation with BraTumIA in Patients with Recurrent Glioblastoma.AJNR Am J Neuroradiol 42, no. 6 (June 2021): 1080–86. https://doi.org/10.3174/ajnr.A7071.
Abu Khalaf N, Desjardins A, Vredenburgh JJ, Barboriak DP. Repeatability of Automated Image Segmentation with BraTumIA in Patients with Recurrent Glioblastoma. AJNR Am J Neuroradiol. 2021 Jun;42(6):1080–6.
Abu Khalaf, N., et al. “Repeatability of Automated Image Segmentation with BraTumIA in Patients with Recurrent Glioblastoma.AJNR Am J Neuroradiol, vol. 42, no. 6, June 2021, pp. 1080–86. Pubmed, doi:10.3174/ajnr.A7071.
Abu Khalaf N, Desjardins A, Vredenburgh JJ, Barboriak DP. Repeatability of Automated Image Segmentation with BraTumIA in Patients with Recurrent Glioblastoma. AJNR Am J Neuroradiol. 2021 Jun;42(6):1080–1086.

Published In

AJNR Am J Neuroradiol

DOI

EISSN

1936-959X

Publication Date

June 2021

Volume

42

Issue

6

Start / End Page

1080 / 1086

Location

United States

Related Subject Headings

  • Tumor Burden
  • Nuclear Medicine & Medical Imaging
  • Magnetic Resonance Imaging
  • Image Processing, Computer-Assisted
  • Humans
  • Glioblastoma
  • Brain Neoplasms
  • Algorithms
  • 3406 Physical chemistry
  • 3209 Neurosciences