Emotion and Development Branch Phenotyping and DTI (2012-2017)

# Emotion and Development Branch Phenotyping and DTI (2012-2017)

A comprehensive dataset characterizing a transdiagnostic sample of youth aged 7 to 20 years using clinical assessments, structural magnetic resonance imaging (MRI), and diffusion tensor imaging (DTI).

Data was collected at the National Institute of Mental Health (NIMH), Bethesda, Maryland by the [Emotion and Development Branch (EDB)](https://www.nimh.nih.gov/research/research-conducted-at-nimh/research-areas/clinics-and-labs/edb). The basic pre-processing scripts are shared on the [study GitHub repository](https://github.com/nimh-dsst/DTI-Phenotyping). This dataset will allow a wide array of investigations into brain structure and psychopathology in youth.

This dataset is licensed under the [Creative Commons Zero (CC0) v1.0 License](https://creativecommons.org/publicdomain/zero/1.0/).

## Participant Eligibility and Characteristics

### Inclusion Criteria

Participants in the study met the following inclusion criteria:

* Aged 7 to 20 years and spoke English
* Met the criteria for attention-deficit/hyperactivity disorder (ADHD) or disruptive mood dysregulation disorder (DMDD) as assessed with the KSADS.
* Did NOT meet the criteria for any psychiatric disorder (i.e., healthy volunteer, HV) as assessed with the KSADS

### Exclusion Criteria

Participants meeting any of the criteria listed below were excluded from the study:

* Neurological disorders
* Pervasive developmental disorders (e.g., autism spectrum disorder)
* Psychosis
* Bipolar disorders
* Substance use
* Any medical condition that increases risk for MRI (e.g., pacemaker, metallic foreign body in eye, dental braces)
* MRI contraindications (e.g., claustrophobia, pregnancy)
* Full-scale IQ < 70

### Consent

Institutional Review Board approval, parent or guardian consent, and child assent were obtained. To characterize the sample, we collected data on participants' race/ethnicity, parental income, and use of psychotropic medication.

## Clinical Measures

### Completed by Participants :
* [Affective Reactivity Index (youth-ARI)](https://doi.org/10.1111%2Fj.1469-7610.2012.02561.x)
* Wechsler Abbreviated Scale of Intelligence (WASI)

### Completed by Participants' parents/caregivers:
* [Affective Reactivity Index (parent-ARI)](https://doi.org/10.1111%2Fj.1469-7610.2012.02561.x)
* [Conners Comprehensive Behavioral Rating Scale (CBRS)](https://doi.org/10.1007/978-0-387-79948-3_1534)
* [Screen for Children’s Affective Reactivity–Home (SCAR-H)](https://doi.org/10.1007/s10826-006-9120-3)

### Completed by Clinicians:
* [Kiddie Schedule for Affective Disorders (KSADS)](https://doi.org/10.1097/00004583-199707000-00021)
* [Children's Global Assessment Scale (CGAS)](https://doi.org/10.1001/archpsyc.1983.01790100074010)

## Imaging Protocol

### Diffusion Weighted Scans
Acquired with a single-shot echo planar imaging sequence (minimal TR = 6751 ms, TE = 79 ms, 62 axial slices, 2.5 mm slice thickness, FoV = 240 mm², matrix size = 128 x 128). Depending on the participant's weight and the voltage, set by the scanner for excitation, the radio frequency pulses extended up to 7.5 seconds in six participants.

All data were acquired with isotropic voxels of 2.5 mm³, but images were reconstructed by the scanner with a slightly higher in-plane resolution for some participants. The default resolution (128 x 128: n = 126) is noted in filenames with `rec-default`. For the remaining six participants (see `participant_id` list below), this resolution was 256 x 256, which is noted in the filenames with `rec-alternate`.

* 22196
* 23067
* 23111
* 23122
* 23393
* 23460

We obtained five unweighted images and 75 images with diffusion gradients (6: b = 300 s/mm², 69: b = 1000 s/mm²) performed across four runs (or five runs where applicable).

### T2-weighted Scans

Acquired with a fast spin-echo sequence/fat suppression. For most participants TR = 7500 ms, TE = 100 ms, 100 axial slices, no gap, FoV = 240 × 192 mm, matrix size = 256 × 192. Depending on the participant's head size, there were different numbers of slices acquired, although this was 100 slices for most participants in the sample.

The default slice thickness (1.7 mm: n = 98) is noted in the filenames with `acq-default`. For some participants, the slice thickness was 2.5 mm (n = 33; see `participant_id` list below), which is noted in the filenames with `acq-alternate01`.

* 20209
* 20255
* 20587
* 20631
* 20968
* 21651
* 22128
* 22166
* 22183
* 22281
* 22421
* 22437
* 22676
* 22686
* 22687
* 22721
* 22803
* 22809
* 22853
* 22892
* 22954
* 23067
* 23111
* 23112
* 23166
* 23195
* 23216
* 23218
* 23284
* 23378
* 23393
* 23401
* 23414

Additionally, for the slice thickness was 3.0 mm one participant (see `participant_id` below), which is noted in the filenames with `acq-alternate02`.

* 20457

### T1-weighted Scans

Acquired for most participants with TR = 7.6 ms, TE = 3.5 ms, no gap, matrix size = 256 × 256. Depending on the participant's head size, there were different numbers of slices acquired, although this was 176 slices for most participants in the sample.

The dslice thickness = 1 mm, For two participants, the scan was acquired at a resolution of 0.86mm x 0.86mm x 1.20mm.

The default resolution (1.00mm x 1.00mm x 1.00mm: n = 130) is noted in the filenames with `acq-default`. For two participants (see `participant_id` list below), the resolution was 0.86mm x 0.86mm x 1.20mm, which is noted in the filenames with `acq-alternate`.

* 21068
* 21785

## Description of Data Formats and Preparation

### Clinical Measures Data

Relevant clinical measures can be found in the `phenotype.tsv` file, with each measure described in the `phenotype.json` file. The `n/a` values indicate that the question may have been skipped over by the participant or not presented to/asked of the participant.

MPlus scripts with sample analysis steps (e.g., factor analyses) and R scripts (for figure generation) can be found in the [study GitHub repository](https://github.com/nimh-dsst/DTI-Phenotyping).

### Imaging Data

Imaging data in this study is distributed in NifTI format. The images were acquired with a 3.0 Tesla field strength and were oriented in the axial plane. Visual inspection was performed to exclude any images with significant artifacts or distortions.

We have de-identified the anatomical scans by defacing the scans using v1.0.0 of the [DSST Defacing Pipeline](https://github.com/nih-fmrif/dsst-defacing-pipeline/).

The scanner outputted incorrect bval and bvec values associated with the original dicom DTI files. Therefore, to increase usability of these data, we have created concatenated NifTI files for the runs accompanied by corrected bval and bvec files using [TORTOISE](https://tortoise.nibib.nih.gov). For participants with five runs but for whom the original directions were of sufficient quality (n = 4, see `participant_id` list below), the respective images were retained and the associated NifTI files and bval and bvec files representing all runs have differing dimensions from the rest of the sample.

* 01090
* 20569
* 20771
* 22888

Information about the DTI directions removed and example analysis scripts can be found in the [study GitHub repository](https://github.com/nimh-dsst/DTI-Phenotyping).