Dyslexia and Spelling Deficit Reading fMRI (ds003126)

This OpenNeuro dataset holds the imaging behind a study of how children with dyslexia and children with a spelling deficit alone process written words. Children read single German words and pseudohomophones aloud in the scanner; pseudohomophones are misspellings that sound like a real word when read. The data were first released in September 2020; snapshot 1.4.0 (September 2026) is published under a CC0 waiver.

Composition

The release contains 58 children aged 8 to 10 years, 24 girls and 34 boys. Twenty-two have typical reading and spelling skills, 16 have an isolated spelling deficit and 20 have dyslexia; the README lists the subject codes of each group and participants.tsv gives group, age and sex. Each child has one T1-weighted image and three task fMRI runs, for 232 image files. The BIDS tree stores the structural image under ses-1 and the functional runs under ses-2.

Acquisition

Scans were made on a Siemens Skyra 3T system with a 20-channel head and neck coil at the Graz site named in the sidecars. Structural images are 1 mm MPRAGE. Functional runs are 2D gradient-echo EPI with a 64 x 64 matrix, 3 mm slices, TR 2.34 s and TE 33 ms. Each run shows 40 items for 3 s, half words and half pseudohomophones, separated by a jittered fixation of 2 to 6 s. Runs hold 122 to 130 volumes because the scanner was stopped by hand.

Annotations

There are no image annotations. Events files give onset, duration, condition, the shown letter string, item number, length and which of two pseudorandomised task versions the child received. A HED annotation of the events is included. Spoken responses were recorded as audio but are not scored, so no accuracy or reaction times are present.

Known limitations

• Events files up to version 1.3.1 mixed two task schedules and should not be used; 1.4.0 rebuilt them from the stimulus logs.
• Six runs have no events file because their timing could not be recovered, and one of them was acquired apart from its paradigm and should be dropped.
• The sample is small and from a single site.