Objective Morphological Classification of Neocortical Pyramidal Cells

Abstract

A consensus on the number of morphologically different types of pyramidal cells (PCs) in the neocortex has not yet been reached, despite over a century of anatomical studies, due to the lack of agreement on the subjective classifications of neuron types, which is based on expert analyses of neuronal morphologies. Even for neurons that are visually distinguishable, there is no common ground to consistently define morphological types. The objective classification of PCs can be achieved with methods from algebraic topology, and the dendritic arborization is sufficient for the reliable identification of distinct types of cortical PCs. Therefore, we objectively identify 17 types of PCs in the rat somatosensory cortex. In addition, we provide a solution to the challenging problem of whether 2 similar neurons belong to different types or to a continuum of the same type. Our topological classification does not require expert input, is stable, and helps settle the long-standing debate on whether cell-types are discrete or continuous morphological variations of each other.

A systematic evaluation of interneuron morphology representations for cell type discrimination

“…For example, they failed to distinguish double-bouquet cells and Martinotti cells in layer 4. Various forms of persistence-based measures performed consistently worse than density maps, at least for the interneurons studied here (originally, persistence diagrams were developed for cortical pyramidal neurons (Adams et al, 2017;Kanari et al, 2019;). We did not extensively evaluate combinations of feature representations, as combining even the best representations did not dramatically improve performance, possible as a result of overfitting to a data set of limited size (see Figure 5).…”

“…Finally, we used persistence images, a recently introduced quantification of neural morphology based on topological ideas Kanari et al, 2018Kanari et al, , 2019. We used four different distance functions (also called filter functions) to construct one-and two-dimensional persistence images, resulting in eight different persistence representations.…”

“…Morphometric statistics have been used in a wide variety of studies across species and brain areas (Lu et al, 2013;Polavaram et al, 2014;Lu et al, 2015;Gouwens et al, 2018) and have been shown to perform well in a one-vs-rest classification of cortical neurons (Mihaljević et al, 2015(Mihaljević et al, , 2018. Persistence, in turn, has lately been shown to distinguish pyramidal neuron types in juvenile rat somatosensory cortex (Kanari et al, 2019). These studies, however, did not directly compare different morphology representations, but rather focused on comparing classification schemes or establishing cell type related differences within their chosen morphological representation.…”

“…Euclidean distance is undefined for persistence diagrams themselves. To circumvent this problem we converted each persistence diagram into a 1D or 2D image and used Euclidean distance on the results as this procedure has been shown to work well in the neural domain (Kanari et al, 2019;.…”

Morphological pseudotime ordering and fate mapping reveals diversification of cerebellar inhibitory interneurons

“…This approach requires sparse labeling of genetically or lineage-related cell types, which is facilitated by transgenic drivers and viral vectors (Harris et al 2014; He et al 2016). Paired with advancements in statistical and computational methods, neuronal subtypes can be effectively parsed using single neuron anatomy, as illustrated for sensory afferents in the skin (Wu et al 2012), olfactory bulb neurons (Tavakoli et al 2018), and pyramidal and GABAergic interneurons in the cortex ; Kanari et al 2019; Gouwens et al 2019). Drawing fine divisions between subtypes can be challenging however, as heterogeneity and continuous variation within cell types are commonly observed (Cembrowski and Scala et al 2020).…”