Introduction: Neurons in the human dentate nucleus are classified into four types according to their morphology (1) and into two types according to their topology (2). Thus this study have two major aims: i) verify or improve previous classification and ii) investigate whether border neurons express the same features or they belong to a different morphological types (3). Material and Methods: Fifteen parameters quantifying four aspects of neuron morphology (surface area and shape of whole neuron, dendritic length and branching complexity) were measured (1). Classification scheme was investigated using neural networks and multidimensional approach (3). Results: The use of neural network didn’t confirm the previous classification on central and border cells, but it showed four neuronal types, based on soma area and dendritic length. Further analysis showed significant differences between two types of border neurons, mainly in parameters which quantify dendritic branching complexity and dendritic length. All methodological approaches demonstrated slight clustering of data: cluster analysis showed two data clusters and separate unifactor analysis indicated inter-cluster differences. Discriminant and correlation-comparison analysis further proved and explained the result on a more cohesive manner. Conclusion: Human dentate nucleus neurons can be classified into four neuron types, according to their quantitative properties. Border neurons can be divided into two different topological types. The obtained neuronal differences were discussed further in relation to the structure and function of the cerebellar network.

(1) Grbatinić I, Marić DL, Milošević NT. Neurons from the adult human dentate nucleus: neural networks in the neuron classification. J Theor Biol. 2015; 370: 11-20. (2) Marić D. Qualitative and quantitative analysis of adult human dentate nucleus neurons morphology (Ph.D. thesis). Medical faculty, University of Novi Sad, Serbia, Balkans, 2010. (3) Grbatinić I, Milošević NT. Classification of adult human dentate nucleus border neurons: artificial neural networks and multidimensional approach. J Theor Biol. 2016; 404: 273-284.

Financing: Instituto Technologico de Santo Domingo (INTEC), Santo Domingo, Republica Dominicana