Automated and representative fascicle selection for computer-assisted morphometry of myelinated nerve fibres in peripheral nerves.

We describe an algorithm that permits representative fascicle selection for sample acquisition from the entire cross-section of a nerve. We calculate the number of fascicles necessary for the acquisition of a given sample volume of nerve fibres from the image magnification, determine the scheme of intrafascicular sampling, and measure the (estimated) nerve fibre density. We start by storing the positions of all fascicles as coordinates of the microscope motor stage. We then recruit the needed fascicles from the file by calling them up out of the storage order with a constant interval, using the quotient of the total number of fascicles and the number of fascicles necessary to obtain the sample size for this purpose. The results obtained on the analysis of 40 specimens of sural and tibial nerves by means of the image analysis system IBAS are reported.