Ensemble segmentation for GBM brain tumors on MR images using confidence‐based averaging. Issue 9 (20th August 2013)

• Record Type: Journal Article
• Authors: Huo, Jing; Okada, Kazunori; van Rikxoort, Eva M.; Kim, Hyun J.; Alger, Jeffry R.; Pope, Whitney B.; Goldin, Jonathan G.; Brown, Matthew S.
• Is Part Of: Medical physics. Volume 40:Issue 9(2013)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2013-08-20
• Subjects: MRI: anatomic, functional, spectral, diffusion -- Segmentation -- Clinical applications; biomedical MRI -- brain -- image classification -- image segmentation -- medical image processing -- tumours; GBM -- ensemble; Involving electronic [emr] or nuclear [nmr] magnetic resonance, e.g. magnetic resonance imaging -- Digital computing or data processing equipment or methods, specially adapted for specific applications -- Image data processing or generation, in general; Cancer -- Medical imaging -- Brain -- Medical image segmentation -- Medical magnetic resonance imaging -- Cluster analysis -- Cellular automata -- Tissues -- Medical image contrast -- Magnetic resonance; Medical physics -- Periodicals; Medical physics; Geneeskunde; Natuurkunde; Toepassingen; Biophysics; Periodicals; Periodicals; Electronic journals; 610.153
• DOI: 10.1118/1.4817475
• Languages: English
• ISSNs: 0094-2405

Abstract

Abstract : Purpose: : Ensemble segmentation methods combine the segmentation results of individual methods into a final one, with the goal of achieving greater robustness and accuracy. The goal of this study was to develop an ensemble segmentation framework for glioblastoma multiforme tumors on single‐channel T1w postcontrast magnetic resonance images. Methods: : Three base methods were evaluated in the framework: fuzzy connectedness, GrowCut, and voxel classification using support vector machine. A confidence map averaging (CMA) method was used as the ensemble rule. Results: : The performance is evaluated on a comprehensive dataset of 46 cases including different tumor appearances. The accuracy of the segmentation result was evaluated using the F 1 ‐measure between the semiautomated segmentation result and the ground truth. Conclusions: : The results showed that the CMA ensemble result statistically approximates the best segmentation result of all the base methods for each case.