From 822b528e31d59fe8a29bff55cb4ee9a749ef18ed Mon Sep 17 00:00:00 2001 From: James Kent Date: Thu, 2 Nov 2023 13:28:56 -0500 Subject: [PATCH 1/2] spelling --- docs/introduction/ecosystem.md | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/docs/introduction/ecosystem.md b/docs/introduction/ecosystem.md index 802beb8..2676aac 100644 --- a/docs/introduction/ecosystem.md +++ b/docs/introduction/ecosystem.md @@ -20,17 +20,17 @@ large scale meta-analysis to provide new insights into the literature, overcomin with sheer scale. With regular updates, Neurosynth was able to keep up with the growth of the literature. The database was released with a permissive license, and accompanied by a Python package to manipulate and analyze it. -Although this approach was surprsingly successful, there were several major limitations to Neurosynth 1.0: +Although this approach was surprisingly successful, there were several major limitations to Neurosynth 1.0: * Meta-analyses were limited by **concepts that can be inferred from large scale text mining** (i.e. frequency of terms in the text). -Although these features proved to be surprsingly useful for well-powered and broad cognitive constructs (e.g. 'emotion'), Neurosynth was not able +Although these features proved to be surprisingly useful for well-powered and broad cognitive constructs (e.g. 'emotion'), Neurosynth was not able to capture the fine-grained details of the neuroimaging literature, or allow users to define their own grouping of studies. * The database is not curated, and therefore contains many **inaccuracies and incomplete** data at both the study and coordinate level. -Aside from obvious extraction erors, automated coordinate extraction lacks the ability to determine critical information, such as whether the activation is positive or negative. +Aside from obvious extraction errors, automated coordinate extraction lacks the ability to determine critical information, such as whether the activation is positive or negative. In addition, it's not possible to segregate the coordinates into distinct contrast, conditions, or studies without manual curation. -* Coordinate-based analyses are inherently **inferior to image-based** meta-analysis, which is becoming increasingly possible with sharing of unthresholded statisical maps in repositories like [NeuroVault][]. +* Coordinate-based analyses are inherently **inferior to image-based** meta-analysis, which is becoming increasingly possible with sharing of unthresholded statistical maps in repositories like [NeuroVault][]. _Neurosynth Compose_ aims to address these limitations: @@ -122,4 +122,4 @@ it is planned that NeuroVault will focus exclusively on image storage and sharin [PyMARE]: https://pymare.readthedocs.io/en/latest/ [scikit-learn]: https://scikit-learn.org/stable/developers/index.html [Sleuth]: http://www.brainmap.org/software.html#Sleuth -[SPM]: https://www.fil.ion.ucl.ac.uk/spm/ \ No newline at end of file +[SPM]: https://www.fil.ion.ucl.ac.uk/spm/ From 5323da642017132d2e82d106cb6564ba884e2067 Mon Sep 17 00:00:00 2001 From: James Kent Date: Thu, 2 Nov 2023 13:29:47 -0500 Subject: [PATCH 2/2] run a large scale meta-analysis --- docs/introduction/faq.md | 7 +++++ docs/tutorial/large-scale.md | 56 +++++++++++++++++++++++++++++++++++- 2 files changed, 62 insertions(+), 1 deletion(-) diff --git a/docs/introduction/faq.md b/docs/introduction/faq.md index e31d7ca..4550ed2 100644 --- a/docs/introduction/faq.md +++ b/docs/introduction/faq.md @@ -36,3 +36,10 @@ You may also make a Version private if you don't want to share your edits with o Neurosynth 1.0 was an online platform for browsing automatically generated large-scale neuroimaging meta-analyses. However, because all analyses were pre-generated, users were unable to define custom meta-analyses using the Neurosynth database. Instead, Neurosynth 1.0 used text mining techniques to automatically group studies based on the frequency of the terms mentioned in the text. Neurosynth Compose is focused on allowing users to overcome the limitations of automated large-scale meta-analysis, by enabling users to annotate studies, and curate sets of studies amenable for meta-analysis. This way, users can systematically define meta-analyses using their own expertise, while still leveraging the Neurosynth database, and an easy-to-use web-based analysis builder to accelerate the meta-analysis process. You can read more about the relationship between Neurosynth 1.0 and Neurosynth Compose under [Ecosystem](./ecosystem#neurosynth-compose) + + +* What is the difference between importing and ingesting studies? + +Importing studies refers to adding studies for curation, +ingesting studies means that you are adding studies to a StudySet for a meta-analysis. + diff --git a/docs/tutorial/large-scale.md b/docs/tutorial/large-scale.md index 56ce32c..69db062 100644 --- a/docs/tutorial/large-scale.md +++ b/docs/tutorial/large-scale.md @@ -3,4 +3,58 @@ sidebar_label: 'Automated Meta-Analysis' sidebar_position: 3 --- -# Exporatory automated large-scale meta-analysis \ No newline at end of file + +# Outline + +1. Create Project +2. Select Simple Workflow +3. Import from Neurostore +4. Enter search term +5. Promote all studies to Included +6. Ingest +7. Say all studies are annotated and included +8. Select the MKDA-Chi-squared + + +# Create Project + +As with any meta-analysis on Neurosynth Compose you will need to create a project. + +# Simple Workflow + +Since you will not be performing any manual curation of the studies, the Simple workflow +is the best choice for this project. +The PRISMA workflow would be overkill for this project as PRISMA involves checking each +individual study for inclusion, which for a large-scale meta-analysis would be very time consuming. + +# Import from Neurostore + +After selecting the simple workflow, you have several options to import studies. +Importing from neurostore is the simplest way to get started since all the studies selected will have coordinates associated with them. +You will be taken to a search box where you can enter the search term you are interested in, while you are not restricted to the terms in the original neurosynth database, +you are not guaranteed to find enough studies if you enter a term that is +not in the original database. + +# Promote studies + +In an automated meta-analysis you will not be performing any manual curation of the studies, so you can promote all the studies to included using the respective button. + +# Ingest + +With all the studies promoted, you can now ingest them as a studyset for use in a meta-analysis. + + +# Say all studies are annotated and edited + +Since you are not performing any manual curation of the studies, you can say that all the studies are annotated and edited + +# Select the MKDA-Chi-squared + +During the algorithm selection stage choose MKDA-Chi-squared + +# Exploratory automated large-scale meta-analysis + +Like the original Neurosynth, Neurosynth Compose allows users to perform automated meta-analyses using the Neurosynth database. +Finding the analyses for the concepts you are interested in is easy using the search bar. +You can enter combinations of search terms into the search bar +