participant_demographics#
You can see the full contents of this project on GitHub.
participant_demographics#
Annotating information about participants: count, sex, age, and diagnosis.
Papers#
How the papers were obtained#
Typically with pubget.
We recommend invoking pubget with the --query_file option, and storing a copy of the query file in the project’s directory, or including a copy in the README.md.
<description>
Where the full papers are stored#
The full pubget output can be found here on OSF: https://osf.io/kfmdp/
Annotations#
File(s) being annotated:#
/projects/<project_name>/documents/<documents_file1_name>.jsonlcorresponding file in the pubget output:
<pubget_folder_name>/subset_allArticles_labelbuddyData/<documents_file1_name>.jsonl
/projects/<project_name>/documents/<documents_file2_name>.jsonlcorresponding file in the pubget output:
<pubget_folder_name>/subset_allArticles_labelbuddyData/<documents_file2_name>.jsonl
Annotation labels:#
<label1>: <description of label1><label2>: <description of label2>
Labels found in other projects as well:#
<label2>
Instructions for annotators#
<description>
Some analyses of this project’s annotations are shown in this page.
How to annotate#
Annotating demographic information about the participants is more complex than other projects in this repository, because studies typically involve several groups of participants, with diverse structures, and there is some variation in how the relevant information is reported. To annotate a piece of information about a group of participants, we stack several annotations on top of each other. We add annotations that identify the group of participants (eg patients vs controls), then an annotation that contains the information of interest (eg count, min age, etc.). To be easily linked these annotations should be at the exact same positions, which is easy to acheive in labelbuddy by clicking several labels in sequence (or by first selecting an existing annotation and then clicking a new label to add it on top).
We consider that most articles roughly conform to the participant group structure depicted in the tree shown in Fig. 1. The root contains all the participants, which are then divided in patients and healthy controls, each of which may contain several subgroups, and finally each subgroup can contain females and males. Note that for many articles, some of the nodes will be empty – eg studies involving only healthy controls, only one sex, etc.
Fig. 1 The participant group structure and corresponding annotations.#
Annotation example with subgroups#
In Fig. 1 we see the general way of annotating information about participants. We start by describing the most complex case but for most annotations the situation will be simpler.
This video (without sound) illustrates the annotation process that is described below.
The report on the left shows a continuously updated summary of the participants in the current document, it is launched with the scripts/watch_participants.py script as explained here.
Here we annotate the count (20) of a specific subgroup, constituted of:
Patients
Within patients, the schizophrenia subgroup (this article also has an autism spectrum disorder subgroup of patients)
Within the schizophrenia subgroup, the males.
To annotate this information, we select the information we want to annotate and then apply labels, starting from the top of the participants tree (the actual order doesn’t matter, this is just a suggestion). We first click the “patients” label.
Then, as there are several patients subgroups in this article, we need to differentiate the schizophrenia subgroup. We don’t want to add new labels for subgroups, as we would end up with a very long list of labels, most of which are used in few papers (eg “schizophrenia”, “siblings”, “experts”, etc.) Instead, we rely on the extra data input field in labelbuddy. While the “patients” annotation is still selected, we write int its extra data field (on the bottom left of labelbuddy). We enter in there whatever name we want to give to the schizophrenia subgroup, which will act as a local identifier within the current paper. This name is arbitrary and only serves to link the different annotations about that subgroup, here we unoriginally chose “schizophrenia”. Referring it to it again for other annotations will be easy because labelbuddy will propose it in a completion list whenever we are entering extra data for the “patients” label.
Next, we click on the “males” label to create a new annotation, indicating that within the “patients” / “schizophrenia” subgroup, we are looking at the males. Finally, we click on the “count” label to create a new annotation, indicating the type of information contained in our selected text. If needed, we can use the extra data here again – for example if the count was indicated as “twenty” (in English), we would enter in the extra data “20” (the value in numbers), to make it easier to use the annotation later.
So to summarize, the steps are:
Select the group (“patients” or “healthy”)
Enter the subgroup identifier in the “extra data” field (with the help of the completion list if we have seen that subgroup before)
Select the sex (“females” or “males”)
Select the label that indicates the type of information (“count”, “age mean”, etc.)
If necessary add any complementary information in the “extra data” field (eg “20” when the selected text is “twenty”).
When we annotate information about nodes that are higher in the participant group tree, we simply omit the labels that do not apply. For example, if we are annotating the total count of participants (healthy and patients), we simply apply the label “count”, without indicating a group, subgroup or sex. As we see below, when we select the diagnosis, we only indicate the group and subgroup, as the diagnosis applies to both males and females.
Here are all the annotations for article discussed above, PMC8883821:
PMC8883821
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Disease-Specific Contribution of Pulvinar Dysfunction to Impaired Emotion Recognition in Schizophrenia
78 participants
Healthy participants
30 participants, 8 females, 22 males
Age range: 19 – 54 years
Patients autism spectrum disorder
Diagnosis: autism spectrum disorder
20 patients, 4 females, 16 males
Age range: 19 – 43 years
Patients schizophrenia
Diagnosis: schizophrenia
28 patients, 8 females, 20 males
Age range: 21 – 54 years
A simpler example#
When the participant structure of an article is simpler, we can omit any of the labels as long as it does not introduce an ambiguity.
For example, if there is only one group of patients, we do not need to indicate a subgroup.
If the study contains only patients or only healthy participants, we do not need to use the patients or healthy labels.
Which label applies will be inferred from the presence of a diagnosis.
The live report can help check that any information we leave out is being correctly inferred as we annotate.
Below is an example for the article PMC3447931 where only the count is provided, for the patients and for the healthy controls. Note that “diagnosis” implicitly refers to patients, so we can omit the group label here (but it would not be an error to add it).
PMC3447931
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Patterns of Spontaneous Brain Activity in Amyotrophic Lateral Sclerosis: A Resting-State fMRI Study
40 participants
Healthy participants
20 participants
Patients
Diagnosis: amyotrophic lateral sclerosis
20 patients
Participant demographics summaries#
The repository contains utilities to extract summaries about the participant groups from an article’s annotations and display them as shown in this page.
scripts/participants_report.py creates a report for all the articles exported from a given annotator and project.
scripts/watch_participants.py serves a live summary of the participant groups in the document we are currently annotating in labelbuddy.
From the root of the repository you can run it with:
scripts/watch_participants.py projects/participant_demographics/Your_Name.labelbuddy
(If you call it without specifying a file it will pick the most recently modified .labelbuddy file in the projects/ directory.)
It will print the path to a file that you can open in a web browser and that can help to check annotations are correctly interpreted as you create them. If possible, the report will be automatically opened in the default web browser.
See scripts/participants_report.py --help and scripts/watch_participants.py --help for details.
Some more examples#
Below are a few more examples of annotated documents to help annotators get started.
PMC9407088
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Custom 3D fMRI Registration Template Construction Method Based on Time-Series Fusion
2226 participants
Healthy participants abide 1
573 participants
Healthy participants abide 2
593 participants
Patients abide 1
Diagnosis: autism spectrum disorder
539 patients
Patients abide 2
Diagnosis: autism spectrum disorder
521 patients
PMC8785614
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Frontal interhemispheric structural connectivity, attention, and executive function in children with perinatal stroke
83 participants
Healthy participants
31 participants
Patients ais
Diagnosis: arterial ischemic stroke
26 patients
Patients pvi
Diagnosis: periventricular venous infarction
26 patients
PMC9230060
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Changes of Brain Structures and Psychological Characteristics in Predatory, Affective Violent and Nonviolent Offenders
40 participants
Healthy participants
20 participants, 20 males
Patients
Diagnosis: offenders
20 patients, 20 males
PMC9108497
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Feasibility of FreeSurfer Processing for T1-Weighted Brain Images of 5-Year-Olds: Semiautomated Protocol of FinnBrain Neuroimaging Lab
134 participants
Healthy participants
134 participants, 62 females, 72 males
Age range: 5.08 – 5.22 years
Age mean: 5.3 years
PMC8828908
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Default Mode Network Alterations Induced by Childhood Trauma Correlate With Emotional Function and SLC6A4 Expression
216 participants
Healthy participants
216 participants
PMC8782893
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
An examination of the relationships between attention/deficit hyperactivity disorder symptoms and functional connectivity over time
167 participants
Healthy participants
79 participants
Patients
Diagnosis: attention/deficit hyperactivity disorder
88 patients
PMC8752963
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Cortical correlation structure of aperiodic neuronal population activity
110 participants
Healthy participants
110 participants
PMC8978988
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Novel compound heterozygous mutations in a GNE myopathy with congenital thrombocytopenia: A case report and literature review
1 participants
Patients
Diagnosis: GNE myopathy with congenital thrombocytopenia
1 patients, 1 males
PMC8933759
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Interhemispheric functional connectivity asymmetry is distinctly affected in left and right mesial temporal lobe epilepsy
46 participants
Healthy participants
23 participants
Patients LMTLE
Diagnosis: Left Mesial Temporal Lobe Epilepsy
12 patients
Patients RMTLE
Diagnosis: Right Mesial Temporal Lobe Epilepsy
11 patients
PMC9205431
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Inferring excitation-inhibition dynamics using a maximum entropy model unifying brain structure and function
76 participants
Healthy participants APOE
38 participants, 22 females, 16 males
Healthy participants NC
38 participants, 22 females, 16 males
PMC9461104
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Clinical, radiological, and genetic variation in pontocerebellar hypoplasia disorder and our clinical experience
6 participants
Patients
Diagnosis: Pontocerebellar hypoplasia
6 patients
PMC9564100
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Human cerebellum and corticocerebellar connections involved in emotional memory enhancement
1418 participants
Healthy participants
1418 participants, 872 females, 546 males
Age range: 18 – 35 years
Age mean: 22.4 years
PMC9548384
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Wearing a KN95/FFP2 facemask induces subtle yet significant brain functional connectivity modifications restricted to the salience network
23 participants
Healthy participants
23 participants, 23 males
Age mean: 29.9 years
PMC9308181
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Renal phenotypes correlate with genotypes in unrelated individuals with tuberous sclerosis complex in China
173 participants
Patients
Diagnosis: tuberous sclerosis complex
173 patients
PMC8939409
See article on PMC
Labelled by Jerome_Dockes
in participant_demographics
Altered brain network topology related to working memory in internet addiction
47 participants
Healthy participants
23 participants, 13 females, 10 males
Age mean: 21.1 years
Patients
Diagnosis: internet addiction
24 patients, 10 females, 14 males
Age mean: 20.6 years