Artificially enlarged training set in image segmentation

Tuomas Tölli, Juha Koikkalainen, Kirsi Lauerma, Jyrki Lötjönen

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

14 Citations (Scopus)

Abstract

Due to small training sets, statistical shape models constrain often too much the deformation in medical image segmentation. Hence, an artificial enlargement of the training set has been proposed as a solution for the problem. In this paper, the error sources in the statistical shape model based segmentation were analyzed and the optimization processes were improved. The method was evaluated with 3D cardiac MR volume data. The enlargement method based on non-rigid movement produced good results – with 250 artificial modes, the average error for four-chamber model was 2.11 mm when evaluated using 25 subjects.
Original languageEnglish
Title of host publicationMedical Image Computing and Computer-Assisted Intervention – MICCAI 2006
EditorsRasmus Larsen, Mads Nielsen, Jon Sporring
Place of PublicationHeidelberg
PublisherSpringer
Pages75-82
Volume1
ISBN (Electronic)978-3-540-44708-5
ISBN (Print)978-3-540-44707-8
DOIs
Publication statusPublished - 2006
MoE publication typeA4 Article in a conference publication
Event9th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI2006) - Copenhagen, Denmark
Duration: 1 Oct 20066 Oct 2006

Publication series

SeriesLecture Notes in Computer Science
Volume4190
ISSN0302-9743

Conference

Conference9th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI2006)
Country/TerritoryDenmark
CityCopenhagen
Period1/10/066/10/06

Keywords

  • Statistical shape models
  • segmentation
  • MR images

Fingerprint

Dive into the research topics of 'Artificially enlarged training set in image segmentation'. Together they form a unique fingerprint.

Cite this