A comparison of two-dimensional image analysis and design-based stereology for evaluating morphological and anatomical changes in the dopaminergic system of the rodent midbrain

Background. 2D analyses produce systematic errors in quantifying anatomical and morphological features in the brain. Design-based stereology overcomes this limitation by applying probability theory, yet many neuroscience investigators still use 2D analyses. The purpose of this study is to compare 2D analysis with design-based stereology in quantifying differences of morphological and anatomical features between groups.;Methods. Brain tissue samples of three different rodent models were analyzed; chronic MPTP/probenecid PD (MPD) mouse model, alcohol preferring (AP) rat model, and the enriched environment (EE) rat model. 2D analyses and design-based stereology were used to quantify neuronal number, neuronal volume and regional volume. Student's t -test (two-tailed) was used to compare quantitative data.;Results. 2D analyses generated significantly different estimation form design-based stereology in neuronal number and did not find relatively small differences of neuronal number. 2D analysis generated comparable value to design-based stereology in normalized data but not in actual value. 2D estimated accurately regional volume.;Discussion. 2D analyses may be used for rough screening to find a difference of neuronal number and volume but should not for the estimation of actual value. Design-based stereology should be used to estimate neuronal number and volume. Both 2D analyses and design based stereology can be used for the estimation of regional volume.