Avascular, neovascularised, and normally vascularised retinal areas were quantified using Image J software.
Totally, 10 random areas were quantified using an imaging analyzing system (CUE-2, Olympus Opticals, Hamburg, Germany), and the numbers were normalized to retinal capillary area.
Images were processed using Adobe PhotoShop 5.0 or 7.0 and colony areas were quantified using the Image Pro Plus 5 software.
CD31 positive areas were quantified using ImageJ software and all data was expressed as means ± SD of duplicate plugs normalized to the positive control (100% vascularization).
Visceral fat areas were quantified using non-contrast CT scans (conditions: 120 kVp, 150 mA, 3 mm slice thickness, 3 mm reconstruction interval) using a Somatom Sensation 16 (Siemens, Munich, Germany).
VEGF-stained images were taken at the lesion epicenter in axial sections at ×50, and the VEGF-positive areas were quantified using light intensity gain counting (n = 3, each).
NF-H-stained images were taken at the epicenter and 4, 3, 2, 1, and 0.5 mm rostral and caudal to the epicenter in axial sections at ×50 magnification, and the NF-H-positive areas were quantified using light intensity gain counting (n = 5, each).
VEGF-A+ areas were quantified using Image J software.
Muscle fiber areas were quantified using NIH Image J software (Bethesda, Maryland, USA).
Lesion areas were quantified using Image-Pro Plus Software Version 5.1.2 (MediaCybernetics, MD) [18,19].
