An SUVref filter was computed for each individual dataset, and the mean filter size across all repeats for a given reconstruction protocol applied to those datasets for the analysis.
Second, to determine the effectiveness of mixing instances of cyber hate across protected characteristics in improving classification of individual types of hate speech, we drew a random sample from each individual dataset (race, sexuality and disability) and combined the samples into a single dataset for training.
Modeltest was run for each individual dataset.
These were reprocessed uniformly and combined into a single training dataset: First, signal intensity was normalized within each individual dataset using Robust Multi-Array Average (RMA) analysis.
First, for each individual dataset (both ALS and healthy controls), the anatomical T1-weighted scan was placed into standard space and grey and white matter were segmented.
In order to minimize the possibility that there are experimental batch effects that associate with OS within each individual dataset, we performed unsupervised hierarchical clustering in each of the 4 datasets.
The effect of this individual cut-off threshold on the group clustering was examined by repeating the individual clustering stage with 3 different settings of rc, being 0.3, 0.4 and 0.5, resulting in 3 clustermaps for each individual dataset.
First, an independent component analysis (ICA) of the resting-state data was performed and 60 spatio-temporal independent components were extracted from each individual dataset using the MELODIC FSL software (MELODIC v4.0; FMRIB Oxford University, UK).
Secondly, for each individual dataset, cortical thickness at every small region of the cortical surfaces was determined by computing the distance between the computed white matter and grey matter surface reconstructions [48].
These methods were applied to each individual dataset.
