Automated nerve fibre size and myelin sheath measurement using microcomputer-based digital image analysis: theory, method and results.

In either clinical or research settings, manual measurement and counting of myelinated fibres in peripheral nerve is tedious and error-prone, yet fully automatic computerized counting and measuring of fibres fails to count small fibres and eliminate extraneous profiles in the tissue. This article describes an operator-interactive, semiautomated method for quantification of myelinated nerve fibre data using commercially available hardware and software on an inexpensive, yet full-featured image analysis system based on a microcomputer. Software macros automate the acquisition of data from the microscope images and the production of numerical and graphic data, with output to either paper hardcopy or 35 mm colour slides. User control is retained for dynamic thresholding, binary image creation and elimination of artifacts. In addition to generating the classic histogram showing the size distribution of nerve fibres, the thickness and variability of myelin sheaths are also graphically depicted. The method is based on measurement of myelin area and total perimeter, with calculation of equivalent circles and diameters for both axon and nerve fibre. Measured fibre sizes are thus somewhat larger than those resulting from manual methods using the minor diameter of an oval profile, or mean diameter of crenated or irregular profiles. The method allows the rapid measurement and counting of numbers of fibres previously impossible to assess manually, or using digitizing tablets. By increasing the speed and accuracy of data acquisition and processing using widely available microcomputers, the method may allow a better description of peripheral nerve changes in research and clinical settings.