Unbiased estimation of total β-cell number and mean β-cell volume in rodent pancreas

A three-dimensional counting rule and its integral test system, the disector, for obtaining unbiased estimates of the number of arbitrary particles in a specimen is presented. Used in combination with ordinary and recently developed stereological methods unbiased estimates of various mean particle sizes and the variance of particle volume are obtainable on sets of two parallel sections with a known separation. The same principle allows the unbiased estimation of the distribution of individual particle volumes in sets of serial sections.

Unbiased estimates of the total number of neurons and glial cells from central regions of grey matter in human brains are obtained using the disector principle in modifications which are unaffected by the histological processing of paraffin-embedded tissue. Section thickness does not enter into the estimator and need not be known. An analysis of the contributions from all sampling levels to the variance of the estimator of total cell number has been used to design an optimimal sampling scheme. Using this method, a precise and unbiased estimate of the total number of neurons in a defined brain region can be obtained in less than 1 h.