Journal de toxicologie clinique
Libre accès

Abstrait

Multivariate PCA Analysis Combined with Ward’s Clustering for Verification of Psychological Characterization in Visually and Acoustically Social Contexts

Koki Mimura, Yuka Shirakawa, Shun Nakamura and Mamiko Koshiba

The development of the nervous system is a process of gene-environment interaction through prenatal and neonatal stages. The chemical environment may cause mal-development of the social brain network leading to developmental disorders. Thus, we need to develop a new behavioral method which distinguishes the subtle difference in early stage development of sensory-motor, socio-emotional, and cognitive processing in humans and animal models. Here, we have developed an innovative tool for signal processing to structuralize the behavioral elements from a large body of complex information via the multivariate analysis (principal components analysis combined with Ward’s clustering) using an animal model (domestic chicks) and found dynamical differences of combined behavioral parameters over social contexts between socially affiliated chicks and isolated chicks. We prepared two groups each containing six subjects, each group had its peer-social environments regulated, either isolated or grouped, and then their social interaction was tested through a series of social contexts consisting of four serial contexts, 1) firstly isolation, 2) acoustical interaction, 3) acoustical and visual meeting with the reference peers, and lastly 4) isolation again. Results show the juveniles’ social cognition and emotion was visualized as specific clustering structures with active and affective behavioral factors that were seen only in context three with unfamiliar peers visually and acoustically but were not seen in isolation or the social auditory without visual context. These results suggest the usefulness of the combined multivariate analysis under sensory cued-restricted contexts to quantitatively reveal a subtle difference of socio-emotional behavior.