Figure 3
Application of the 2D nucleator on NeuN-stained tissue sections. A shows representative coronal NeuN-stained brain sections developed with nickel-enhanced DAB from outbred male sham SDs (first row) and SHRs (second row). Note the larger ventricular system and the flatter appearance of the brain on the coronal sections of SHRs. B and C are larger images of one of the NeuN stains from an animal with and without ischemic stroke. The cross-sectional area of the infarct (B) or the hemisphere (C) is estimated with the 2D nucleator [21, 23]. Our estimates were based on computer-assisted measurements using CAST® software (Visiopharm A/S, Hørsholm, Denmark). After marking the center in the region of interest, the computer generated four random, right-angled, isotropic directions. The intersection lengths (ℓ+) from the center point to the boundary were measured and the cross sectional area was estimated with the equation: 
. Cross sections of more complex structures included several intercepts from the center point to the boundary (C). ℓ2 was calculated by adding the squared lengths with alternating opposite signs (ℓ+ or ℓ-). The intersection point most distant from the centre was always positive. Finally, the volume of the brain structure or infarct could be estimated using the Cavalieri principle, since: (1) the coronal slices were parallel; (2) the distance between the slices was known (i.e. 600 μm); and (3) the first coronal slice hitting the structure of interest was random. NeuN, neuronal nuclei; SDs, Sprague-Dawley rats; SHRs, spontaneously hypertensive rats.