Avian telencephalon and cerebellum volumes can be accurately estimated from digital brain endocasts.

For studies of the evolution of vertebrate brain anatomy and potentially associated behaviours, reconstructions of digital brain endocasts from computed tomography scans have revolutionized our capacity to collect neuroanatomical data. However, measurements from digital endocasts must be validated as reflecting actual brain anatomy, which is difficult because the collection of soft tissue information through histology is laborious and time-consuming. In birds, the reliability of digital endocast measurements as volume proxies for the two largest brain regions-the telencephalon and cerebellum-remains to be validated despite their use as proxies, e.g. of cognitive performance or flight ability. We here use the largest dataset of histology and digital endocasts to date, including 136 species from 25 avian orders, to compare digital endocast surface area measurements with actual brain volumes of the telencephalon, cerebellum and whole-brain endocast. Using linear and phylogenetically informed regression analyses, we demonstrate that endocast surfaces are strongly correlated with their brain volume counterparts for both absolute and relative size. This provides empirical support for using endocast-derived cerebellar and telencephalic surface areas in existing and future studies of living and extinct birds, with potential to expand to the dinosaur-bird transition in the future.