Robert-Fitzgerald, T. | Search bioRxiv

bioRxiv2022

Curation of BIDS (CuBIDS): a workflow and software package for streamlining reproducible curation of large BIDS datasets

The Brain Imaging Data Structure (BIDS) is a specification accompanied by a software ecosystem that was designed to create reproducible and automated workflows for processing neuroimaging data. BIDS Apps flexibly build workflows based on the metadata detected in a dataset. However, even BIDS valid metadata can include incorrect values or omissions that result in inconsistent processing across sessions. Additionally, in large-scale, heterogeneous neuroimaging datasets, hidden variability in metadata is difficult to detect and classify. To address these challenges, we created a Python-based software package titled "Curation of BIDS" (CuBIDS), which provides an intuitive workflow that helps users validate and manage the curation of their neuroimaging datasets. CuBIDS includes a robust implementation of BIDS validation that scales to large samples and incorporates DataLad--a version control software package for data--to ensure reproducibility and provenance tracking throughout the entire curation process. CuBIDS provides tools to help users perform quality control on their images metadata and identify unique combinations of imaging parameters. Users can then execute BIDS Apps on a subset of participants that represent the full range of acquisition parameters that are present, accelerating pipeline testing on large datasets. HIGHLIGHTSO_LICuBIDS is a workflow and software package for curating BIDS data. C_LIO_LICuBIDS summarizes the heterogeneity in a BIDS dataset. C_LIO_LICuBIDS prepares BIDS data for successful preprocessing pipeline runs. C_LIO_LICuBIDS helps users perform metadata-based quality control. C_LI

neuroscience↗

bioRxiv2020

A simple permutation-based test of intermodal correspondence

Many key findings in neuroimaging studies involve similarities between brain maps, but statistical methods used to measure these findings have varied. Current state-of-the-art methods involve comparing observed group-level brain maps (after averaging intensities at each image location across multiple subjects) against spatial null models of these group-level maps. However, these methods typically make strong and potentially unrealistic statistical assumptions, such as covariance stationarity. To address these issues, in this paper we propose using subject-level data and a classical permutation testing framework to test and assess similarities between brain maps. Our method is comparable to traditional permutation tests in that it involves randomly permuting subjects to generate a null distribution of intermodal correspondence statistics, which we compare to an observed statistic to estimate a p-value. We apply and compare our method in simulated and real neuroimaging data from the Philadelphia Neurodevelopmental Cohort. We show that our method performs well for detecting relationships between modalities known to be strongly related (cortical thickness and sulcal depth), and it is conservative when an association would not be expected (cortical thickness and activation on the n-back working memory task). Notably, our method is the most flexible and reliable for localizing intermodal relationships within subregions of the brain and allows for generalizable statistical inference.

neuroscience↗