• DocumentCode
    720204
  • Title

    Digital pathology: Identifying spongiosis in unstained histopathology specimen

  • Author

    Abeysekera, Sanush ; Kar, Nicholas ; Wei Siew ; Po-Leen Ooi, Melanie ; Ye Chow Kuang ; Syed Hassan, Sharifah ; Demidenko, Serge

  • Author_Institution
    Sch. of Eng., Monash Univ. Malaysia, Malaysia
  • fYear
    2015
  • fDate
    11-14 May 2015
  • Firstpage
    1970
  • Lastpage
    1975
  • Abstract
    Histopathological specimens are prepped through a process called staining prior to analysis by the pathologist. Staining of a pathological specimen is a standard procedure used to increase the contrast between the cell and tissue structures against the background. Unfortunately, staining is a lengthy process that requires hours of preparation. Moreover, the chemicals used to perform the procedure can affect the specimen´s characteristics. The entire problem of staining can be eliminated if detection and diagnosis can be performed on unstained specimen. However the low-contrast unstained samples can seriously affect the diagnosis reliability. Currently, no established technique exists for the detection and diagnosis of unstained histhopathological samples. This project aims to detect and diagnose spongiosis, a type of cerebral edema, in unstained histopathological samples taken from poultry brains. Success of this research is the first step towards detecting various classes of cerebral edema. It is a fast and accurate clinical tool that greatly enhances the analytic capability of the histopathological laboratory. The computer-aided diagnosis enables short turn-around time and higher consistency in the histopathological laboratory that services hospitals and clinics.
  • Keywords
    biological specimen preparation; biological tissues; biomedical optical imaging; brain; cellular biophysics; diseases; feature extraction; medical diagnostic computing; neurophysiology; cell structure contrast; cerebral edema class; cerebral edema detection; clinical tool; computer-aided diagnosis; digital pathology; histopathological laboratory analytic capability; histopathological laboratory consistency; histopathological specimen preparation; histopathological specimen staining; hospital; low-contrast unstained sample effect; poultry brain; short turn-around time; specimen characteristics; spongiosis detection; spongiosis diagnosis reliability; spongiosis identification; staining chemical effect; tissue structure contrast; unstained histopathological specimen; Accuracy; Algorithm design and analysis; Classification algorithms; Feature extraction; Multispectral imaging; Pathology; Standards; digital pathology; machine learning; spectral imaging;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Instrumentation and Measurement Technology Conference (I2MTC), 2015 IEEE International
  • Conference_Location
    Pisa
  • Type

    conf

  • DOI
    10.1109/I2MTC.2015.7151584
  • Filename
    7151584