Premium
Image analysis of Nissl‐stained neuronal perikarya in the primary visual cortex of the rat: Automatic detection and segmentation of neuronal profiles with nuclei and nucleoli *
Author(s) -
Ahrens Peter,
Schleicher Axel,
Zilles Karl,
Werner Lisa
Publication year - 1990
Publication title -
journal of microscopy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.569
H-Index - 111
eISSN - 1365-2818
pISSN - 0022-2720
DOI - 10.1111/j.1365-2818.1990.tb02970.x
Subject(s) - nissl body , visual cortex , biology , golgi apparatus , segmentation , nucleus , cytoarchitecture , nucleolus , cortex (anatomy) , morphological analysis , anatomy , neuron , neuroscience , staining , artificial intelligence , microbiology and biotechnology , computer science , genetics , endoplasmic reticulum
SUMMARY An image analysing procedure for the morphometric characterization of cortical neurons in Nissl‐stained brain sections is described. It consists of the automatic detection of cellular profiles and their compartments: cytoplasm, nucleus and nucleolus. The algorithm was designed to cope with the large morphological spectrum of cortical perikarya (e.g. geometrical properties of perikarya, staining intensities of cell compartments and nucleo‐plasmic area‐ratio) including pyramidal (Golgi‐category I) and non‐pyramidal (Golgi‐category II) neurons. Clusters of cells were separated and non‐neuronal structures (e.g. glia, endothelial cells) as well as tangential, non‐nucleolated sections through neuronal perikarya recognized and excluded from further analysis without requiring interactive procedures. The performance of the profile recognition procedure was evaluated using 426 nucleolated and non‐nucleolated profiles of different types of neurons in the primary visual cortex of the rat. Nucleolated profiles were recognized as such with a 91% accuracy, non‐nucleolated profiles were rejected correctly in 90% of cases. After automatic segmentation and selection of nucleolated neuronal profiles from the microscopic field, a large set of quantitative morphological features including geometrical, densitometrical and textural parameters can be measured using high power light microscopy. This permits quantitative morphometric characterization of different neuronal types. This procedure is the first part of a system for the automatic classification of Nissl‐stained cortical neurons.