MORPHOMETRIC ANALYSIS OF HUMAN CORNEAL ENDOTHELIUM BY MEANS OF SPATIAL POINT PATTERNS
Authors:
- 1 March 2002
- journal article
- research article
- Published by World Scientific Pub Co Pte Ltd in International Journal of Pattern Recognition and Artificial Intelligence
- Vol. 16 (02) , 127-143
Abstract
This paper presents a method for detecting abnormalities in spatial arrangements of cells within any tissue that can be described by different sets of relevant points. The method has been applied to the detection of subtle abnormalities in corneal endothelia. Images of this type of tissue can be characterized by two types of points: cell centroids and triple points associated with the apical intersections as it was proposed by Díaz.7 Both types of points jointly considered are modeled using a bivariate spatial point process; then a statistical analysis based on certain distributional descriptors proposed by Doguwa4,9 is carried out to discriminate severe and subtle abnormalities from controls. According to the results, the method constitutes a valid alternative to currently used techniques, which involve statistical descriptions of cell density, cell area and hexagonality, that fail in the detection of these subtle abnormalities.Keywords
This publication has 7 references indexed in Scilit:
- Testing abnormality in the spatial arrangement of cells in the corneal endothelium using spatial point processesStatistics in Medicine, 2001
- Granulometric moments and corneal endothelium statusPattern Recognition, 2001
- Measuring the Spatial Homogeneity in Corneal Endotheliums by Means of a Randomization TestLecture Notes in Computer Science, 1999
- Fourier Transform Analysis of Human Corneal Endothelial Specular PhotomicrographsExperimental Eye Research, 1997
- Gibbs random fields, cooccurrences, and texture modelingIEEE Transactions on Pattern Analysis and Machine Intelligence, 1994
- On Edge Corrections for the Point‐Event Analogue of the Clark‐Evans StatisticBiometrical Journal, 1988
- A Method for Obtaining Skeletons Using a Quasi-Euclidean DistanceJournal of the ACM, 1968