M. Hilliges et O. Johansson, Comparative analysis of numerical estimation methods of epithelial nerve fibers using tissue sections, J PERIPH N, 4(1), 1999, pp. 53-57
The proper assessment of neuron numbers in the nervous system during physio
logical and pathological conditions, as well as following various treatment
s, has always been an important part of neuroscience. The present paper eva
luates three methods for numerical estimates of nerves in epithelium: I) un
biased nerve fiber profile and nerve fiber fragment estimation methods, II)
the traditional method of counting whole nerve fibers, and III) the nerve
fiber estimation method. In addition, an unbiased nerve length estimation m
ethod was evaluated. Of these four methods, the nerve length per volume met
hod was theoretically optimal, but more time-consuming than the others. The
numbers obtained with the methods of nerve fiber profile, nerve fragment a
nd nerve fiber estimation are dependent on the thickness of the epithelium
and the sections as well as certain shape factors of the counted fiber. How
ever for those, the actual counting can readily be performed in the microsc
ope and is consequently quick and relatively inexpensive. The statistical a
nalysis showed a very good correlation (R > 0.96) between the three numeric
al methods, meaning that basically any method could be used. However, depen
dent on theoretical and practical considerations and the correlation statis
tics, it may be concluded that the nerve fiber profile or fragment estimati
on methods should be employed if differences in epithelial and section thic
kness and the nerve fibers shape factors can be controlled. Such drawbacks
are not inherent in the nerve length estimation method and, thus, it can ge
nerally be applied.