• DocumentCode
    2948012
  • Title

    Multimodal method for cell membrane extraction in hepatic histological images

  • Author

    Matsushita, Nobumitsu ; Takahashi, Masanobu ; Nakano, Masayuki

  • Author_Institution
    Grad. Sch. of Eng., Shibaura Inst. of Technol., Saitama, Japan
  • fYear
    2010
  • fDate
    Aug. 31 2010-Sept. 4 2010
  • Firstpage
    4068
  • Lastpage
    4071
  • Abstract
    A multimodal method, which uses different kinds of imaging methods, was applied to extract cell membranes in microscopic images of hematoxylin and eosin-stained hepatic histological sections. Cell membrane extraction in hepatic histological images is difficult because the color difference between the cell membrane and the cytoplasm is small in bright-field images. Three kinds of imaging methods, bright-field, dark-field, and phase-contrast imaging, were used because they are readily available for general pathologists. These imaging methods can be switched easily by revolving a combined condenser with the same phase-contrast objective lens. Therefore, little additional time and cost are needed for this approach. Experimental results show the effectiveness of this approach. The correct rate was improved by using additional color information obtained by dark-field and phase-contrast images compared to conventional bright-field images. The best correct rate was obtained when color information of all three images was used. A graphical user interface to calculate the N/C ratio was developed by combining cell membrane extraction with conventional cell nucleus extraction.
  • Keywords
    biomedical optical imaging; biomembranes; feature extraction; graphical user interfaces; image colour analysis; lenses; liver; medical image processing; bright-field imaging; cell membranes; color information; condenser; dark-field imaging; eosin-stained hepatic histological sections; graphical user interface; hematoxylin-stained hepatic histological sections; multimodal method; phase-contrast imaging; phase-contrast objective lens; Artificial neural networks; Biomembranes; Graphical user interfaces; Image color analysis; Microscopy; Support vector machines; Animals; Cell Membrane; Liver; Mice; ROC Curve;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Engineering in Medicine and Biology Society (EMBC), 2010 Annual International Conference of the IEEE
  • Conference_Location
    Buenos Aires
  • ISSN
    1557-170X
  • Print_ISBN
    978-1-4244-4123-5
  • Type

    conf

  • DOI
    10.1109/IEMBS.2010.5627613
  • Filename
    5627613