Pypes: Workflows for Processing Multimodal Neuroimaging Data
Every year, enormous amounts of scientific data are made available to the public (Poline et al., 2012). This trend is due to an increasing demand for transparency, efficiency, and reproducibility. Neuroimaging is a salient example of this trend.In response to the growing concern about the need of pu...
Uložené v:
| Vydané v: | Frontiers in neuroinformatics Ročník 11; s. 25 |
|---|---|
| Hlavní autori: | , , , |
| Médium: | Journal Article |
| Jazyk: | English |
| Vydavateľské údaje: |
Switzerland
Frontiers Research Foundation
11.04.2017
Frontiers Media S.A |
| Predmet: | |
| ISSN: | 1662-5196, 1662-5196 |
| On-line prístup: | Získať plný text |
| Tagy: |
Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
|
| Shrnutí: | Every year, enormous amounts of scientific data are made available to the public (Poline et al., 2012). This trend is due to an increasing demand for transparency, efficiency, and reproducibility. Neuroimaging is a salient example of this trend.In response to the growing concern about the need of publishing relevant software codes (Ince et al., 2012) in the context of results' reproducibility, there is an increasing number of open source initiatives that support code distribution and co-development (Halchenko and Hanke, 2012). The growing diversity of imaging modalities demand from the practitioner a deep technical knowledge of data pre- and post-processing. Consequently, there are open and free tools facilitating image data analysis, e.g., the Python module Nipype1. It offers a homogeneous programming interface and integrates many of these data processing tools. In this sense, resting-state functional magnetic resonance imaging (rsfMRI) is receiving considerable attention by the community with tools such as the Configurable Pipeline for the Analysis of Connectomes (C-PAC)2, and the Data Processing Assistant for Resting-State fMRI (DPARSF)3.As a further contribution to this development, this paper presents a new Python module Pypes—https://github.com/Neurita/pypes. It includes a collection of workflows, reusable neuroimaging pipelines using Nipype, along with some utilities. This library seeks to simplify the reusability and reproducibility of multimodal neuroimaging studies, offering pre- and post-processing utilities inspired by C-PAC. It pre-processes Positron Emission Tomography (PET) and three MRI-based modalities: structural, rsfMRI, and diffusion-tensor MRI (DTI). It also shares an easy-to-use pipeline for COBRE4, a public available dataset. Pypes has been motivated by a need for efficient and reproduceable brain PET/MRI data processing methods. Namely, hybrid PET/MRI scanners become a relevant source of multimodal imaging data, posing new computational challenges. For instance, a simultaneous measurement of brain glucose metabolism and functional connectivity (Aiello et al., 2015; Riedl et al., 2016) opens new perspectives in neuroscience. Structural, functional, and metabolic imaging protocols have been proposed for clinical evaluation of dementia and neuro-oncological cases (Werner et al., 2015; Henriksen et al., 2016). Pypes' immediate motivation was to process PET/MRI data from an ongoing study with more than 400 subjects with suspected neurodegenerative disorders.The paper is organized as follows. After introducing the Python neuroimaging ecosystem and specifically Nipype, we show how to prepare image data for the workflows available in Pypes. Then, we describe worflow configuration for specific imaging modalities. Finally, we present the Pypes pre-processing pipelines and the post-processing utilities. We finish the paper with conclusions and future developments. |
|---|---|
| Bibliografia: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23 Edited by: Juan Manuel Gorriz, University of Granada, Spain Reviewed by: Bogdan Raducanu, Autonomous University of Barcelona, Spain; Michal Wozniak, Wrocław University of Technology, Poland |
| ISSN: | 1662-5196 1662-5196 |
| DOI: | 10.3389/fninf.2017.00025 |