Peripheral nerve morphometry for daily practice.

Quantitative methods for the routine diagnostic analysis of sural nerve biopsies were evaluated. Semiautomatic image analysis was used to compare three sampling methods and to test the accuracy and reproducibility of measurements of various elements in biopsy specimens from seven cases of paraproteinemic neuropathy. Systematic nonrandom sampling gave better results than did completely random sampling, as compared with systematic covering sampling. The accuracy of measurements was optimal at a final magnification of 3,200X. Comparison of single data of double measurements revealed a mean relative frequency of class change of 7.4%. Computation of regression showed good conformity, with r = 1.0. In conclusion, systematic nonrandom sampling at a final magnification of 3,200 X provided good analysis of the total population of structural elements with sufficient accuracy and reproducibility.