Endothelmorphometrie durch automatisierte Fernsehbildanalyse

Abstract

Report on the use of a new image processing system (IPS) for automated morphometry of the corneal endothelium. The method is demonstrated with endothelial photographs taken with the aid of a Keeler-Konan wide angle specular microscope. The original picture is enlarged ten times, then digitized, detected, and analyzed. Areas containing 100-250 endothelial cells can be evaluated automatically. Depending on the quality of the photographs, larger areas sometimes require interaction by the investigator. The evaluation of a 621-cell picture is demonstrated step by step. The time saving which the method offers and its objectivity are discussed, as well as its use in combination with different specular microscopes.

Endotheliale Keratopathie bei Pseudoexfoliationssyndrom: Quantitative und qualitative Morphometrie mittels automatisierter Videobildanalyse

“…Die vorgestellte Methode der automatisierten Endothelmorphometrie ermoglicht die direkte Analyse eines Video-Endothelbildes unter Umgehung einer Filmvorlage. DemgegenUber werden bei den meisten bisher beschriebenen automatisierten Verfahren Photographien oder Videobänder verwendet, um die spekularmikroskopischen Bilder festzuhalten [12,20,40]. Die Auswertung erfolgt dabei erst in einem getrennten zweiten Schritt, nachdem die Aufnahmen in den Computer emgelesen worden sind.…”

“…Das Prinzip der Spekularmikroskopie ist auch Grundlage der heutigen Darstellungs-und Auswertungsverfahren des Hornhautendothels [1, 14,15,17-19, 22, 50, 55]. Man unterscheidet dabei die ,,Non-Contact"-von der ,,Contact"-Methode sowie die Aufnahme mit Normal-und Weitwinkeloptik [21,25,41,43] [12,20,40,51].…”

<title>Morphological image processing and network analysis of cornea endothelial cell images</title>

“…The initial work on the problem of edge detection is represented jy the works of Laing et al [9] and Fabian et al [3]: photographs or video images were digitized and histogram manipulation followed by thresholding was used for edge detection of the cell boudaries. More elegant image processing techniques were recently proposed by Cazuguel et al {2J: histogram equalization, top-hat filter followed by skeleton by influence zones [10], and Yu et al: adaptive local enhancement, low pass filter and matched filters with kernels corresponding to different orientations...…”

Automated video image morphometry of the corneal endothelium

“…To overcome these problems sophisticated computer technology has been developed in order to determine cell boundaries directly from the original photograph or video image, although some manual involvement may still be required [12][13][14][15][16][17][18]. This computerized technology not only quantifies cell densities but also determines the frequency distribution of cell areas and analyzes the polygonality [19].…”