Brain virtual histology of a lizard species (Podarcis bocagei) using X-ray micro-tomography and deep-learning segmentation

Abstract

Lately, there has been an emphasis on the importance of studying inter-individual variation in animal behaviour and cognition and understanding its underlying mechanisms. What was once considered mere noise around population mean can be explained by individual characteristics such as brain morphology and functionality. However, logistical limitations can be faced when studying the brain, especially for research involving wild animals, such as dealing with small sample sizes and time-consuming methods. Here, we combined an efficient and accurate method using X-ray micro-tomography and deep-learning (DL) segmentation to estimate the volume of six main brain areas of wild lizards, Podarcis bocagei: olfactory bulbs, telencephalon, diencephalon, midbrain, cerebellum and brain stem. Through quantitative comparison, we show that a sufficient deep-learning neural network can be trained with as few as five data sets. From this, we applied the trained deep-learning algorithm to obtain volume data of the six brain regions from 29 brains of Podarcis bocagei. We provide a detailed protocol for our methods, including sample preparation, X-ray tomography, and 3D volumetric segmentation. Our work is open-access and freely available, with the potential to benefit researchers in various fields, such as animal physiology, biomedical studies, and computer sciences.