Automatic segmentation of the cerebellum using volBrain software in normal paediatric population

Abstract

AbstractBackgroundMorphological differences that can lead to cerebellar volume changes are associated with the pathogenesis of paediatric diseases. The aim of this study was to examine cerebellum volume in a healthy paediatric population.Materials and methodsTo provide MRI‐based volumetric measurements of the cerebellum, images from the years 2019 to 2021 were scanned retrospectively. A total of 100 images, including the paediatric population aged 0–15 years, were imported into the volBrain software. Volumetric segmentations were obtained automatically, and each lobular cerebellar volume was obtained. The samples were divided into groups of 0–2 years (n = 18), 3–5 years (n = 24), 6–11 years (n = 34) and 12–15 years (n = 24). Obtained cerebellar volumes, age groups, gender and bilateral side comparisons were made.ResultsIn the comparative analyses performed for the total cerebellum and each of the 12 lobular segments, statistically significant differences were found between the age groups in all measurements except Crus II, lobules VIIB, VIIIA and VIIIB (p < 0.05). In multiple comparison tests, statistically significant differences were found between defined age groups, especially infants and toddlers and early adolescence groups (p < 0.05). There was a significant positive correlation between the ages of the subjects and their cerebellum volumes (p < 0.05). Statistically significant differences were found in lobules I–II, VI, VIIIB, IX and X in right and left side volumes (p < 0.05).ConclusionThere is a tendency to increase in cerebellar volume during the transition from childhood to adolescence. The cerebellum has volumetric differences in the first years of life and during adolescence. When the development of a healthy cerebellum is analysed based on volumetric segmentation, differences are observed. The findings of this study may be useful in confirming various theories attributed to the cerebellum in the clinic.