cluster_cst: Clustering Method for the Cortico-Spinal Tract

In astamm/fdatractography: Functional Data Analysis for Tractography in R

Description

This function uses the Mclust method from the mclust package to fit a Gaussian mixture model to the end-points of the CST on the precentral gyrus via Expectation-Maximization for selecting the optimal clustering according to the Bayesian Information Criterion.

Usage

cluster_cst(cst, validate = TRUE)

Arguments

cst	An object of class tract representing the cortico-spinal tract.
validate	A boolean specifying whether the input object should be checked or not (default: TRUE). For this function, it has to be of class tract and represent the cortico-spinal tract (name field of the tract should read "CST").

Value

An object of class Mclust providing the BIC-optimal Gaussian mixture model for clustering the CST streamlines based on cortical position of their end-point.

Examples

file <- system.file("extdata", "Case001_CST_Left.csv", package = "fdatractography")
cst_left <- read_tract(file)
cl <- cluster_cst(cst_left)

astamm/fdatractography documentation built on May 12, 2019, 5:37 a.m.