Comparative Analysis of Unsupervised Algorithms for Breast MRI Lesion Segmentation

  • Sulaiman Vesal
  • Nishant Ravikumar
  • Stephan Ellman
  • Andreas Maier
Conference paper
Part of the Informatik aktuell book series (INFORMAT)


Accurate segmentation of breast lesions is a crucial step in evaluating the characteristics of tumors. However, this is a challenging task, since breast lesions have sophisticated shape, topological structure, and variation in the intensity distribution. In this paper, we evaluated the performance of three unsupervised algorithms for the task of breast Magnetic Resonance (MRI) lesion segmentation, namely, Gaussian Mixture Model clustering, K-means clustering and a markercontrolled Watershed transformation based method. All methods were applied on breast MRI slices following selection of regions of interest (ROIs) by an expert radiologist and evaluated on 106 subjects’ images, which include 59 malignant and 47 benign lesions. Segmentation accuracy was evaluated by comparing our results with ground truth masks, using the Dice similarity coefficient (DSC), Jaccard index (JI), Hausdorff distance and precision-recall metrics. The results indicate that the marker-controlled Watershed transformation outperformed all other algorithms investigated.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Xi X, Shi H, Han L, et al. Breast tumor segmentation with prior knowledge learning. Neurocomputing. 2017;237(Supplement C):145 – 157.Google Scholar
  2. 2.
    Jayender J, Chikarmane S, Jolesz FA, et al. Automatic segmentation of invasive breast carcinomas from dynamic contrast-enhanced MRI using time series analysis. J Mag Res Imaging. 2014;40(2):467–475.Google Scholar
  3. 3.
    Thomassin-Naggara I, Trop I, Lalonde L, et al. Tips and techniques in breast MRI. Diagnost Intervent Imaging. 2012;93(11):828 – 839.Google Scholar
  4. 4.
    Zhang H, Fritts JE, Goldman SA. Image segmentation evaluation: a survey of unsupervised methods. Comput Vis Image Understg. 2008;110(2):260 – 280.Google Scholar
  5. 5.
    Amrehn M, Glasbrenner J, Steidl S, et al. Comparative evaluation of interactive segmentation approaches. In: Bildverarbeitung für die Medizin 2016. Berlin Heidelberg; 2016. p. 68–73.Google Scholar
  6. 6.
    Moftah HM, Azar AT, Al-Shammari ET, et al. Adaptive k-means clustering algorithm for MR breast image segmentation. Neural Comput Appl. 2014 Jun;24(7-8):1917–1928.Google Scholar
  7. 7.
    Soffientini CD, De Bernardi E, Zito F, et al. Background based gaussian mixture model lesion segmentation in PET. Med Phys. 2016;43(5):2662–2675.Google Scholar
  8. 8.
    Vesal S, Diaz-Pinto A, RaviKumar N, et al. Semi-automatic algorithm for breast MRI lesion segmentation using marker-controlled watershed transformation. In: IEEE Nuclear Science Symposium and Medical Imaging Conference Record; 2017. In press.Google Scholar
  9. 9.
    Diaz A, Morales S, Naranjo V, et al. Glaucoma diagnosis by means of optic cup feature analysis in color fundus images. Proc EUSIPCO. 2016 Aug; p. 2055–2059.Google Scholar
  10. 10.
    Reza AM. Realization of the contrast limited adaptive histogram equalization (CLAHE) for real-time image enhancement. J VLSI Sign Process Syst Sign Image Vid Tech. 2004 Aug;38(1):35–44.Google Scholar
  11. 11.
    Xu S, Liu H, Song E. Marker-controlled watershed for lesion segmentation in mammograms. J Digit Imaging. 2011 Oct;24(5):754–763.Google Scholar

Copyright information

© Springer-Verlag GmbH Deutschland 2018

Authors and Affiliations

  • Sulaiman Vesal
    • 1
  • Nishant Ravikumar
    • 1
  • Stephan Ellman
    • 2
  • Andreas Maier
    • 1
  1. 1.Fakultät für Pattern RecognitionFAU Erlangen-NürnbergErlangenDeutschland
  2. 2.Radiologisches InstitutUniversitätsklinikum ErlangenErlangenDeutschland

Personalised recommendations