Precision drawing during task fMRI in healthy adults and individuals with peripheral nerve injury

# README ## Data Access Information - Contact Person Benjamin Philip [e-mail] https://orcid.org/0000-0001-5467-8384 ## Overview - Project Name Precision drawing during task fMRI in healthy adults and individuals with peripheral nerve injury - Years Data were collected April 2022 - June 2025, inclusive. - Task overview The purpose of the study was to identify the neural mechanisms of drawing with the non-dominant left hand, and whether those mechanisms differ between healthy adults and patients with chronic right hand impairment due to peripheral nerve injury. The primary data was task fMRI collected while participants completed a precision drawing task with each hand (3 runs per hand), with accompanying data on drawing performance (30 Hz). Additional scans included resting state (3 runs), t1, t2, and diffusion. Additional behavioral tasks include the Block Building Task (BBT; measures left/right hand choices during unconstrained goal-oriented action) and Box and Blocks, plus 14 assessments of function, activities, and relevant personal history. - Dataset contents 5541 Files, 132.9 GiB 71 Subjects, 0 sessions Available Tasks: drawLH, drawRH, restingstate, blockBuilding Available Modalities: MRI, beh - Independent variables Task: drawLH vs. drawRH - Dependent variables Numerous drawing-performance variables, see participants.json and study-level task-draw*.json. ## Methods ### Subjects Participants: n = 71 right-handed adult humans. Includes 25 adults with chronic unilateral peripheral nerve injury to the right hand/arm (participant id range 1XXX), and 46 healthy adults (participant id range 2XXX). This 71 includes all participants with successfully collected task fMRI data. Two additional participants are not included due to lost data (2027) or incomplete data (2055, stopped after 2/6 task runs because could not use tablet effectively). Of the 71 in this dataset, 5 were excluded from analyses (patients 1020, 1026, 1051; controls 2011, 2019) for reasons listed in participants.tsv and detailed in publications. All participants were right-handed adults with no history of neurological injury. Patients had unilateral peripheral nerve injury to the right hand/arm with at least some difficulty writing. For full inclusion/exclusion criteria, see publications e.g. doi.org/10.1101/2025.11.18.689091. (Note: that study included only the first 33 participants, taken chronologically, due to the study-specific question and power analysis.) ### Apparatus Participants used a MRI-compatible drawing tablet during fMRI scans. ### Task organization The scan order was: T1, drawing task, DTI, T2, restingstate. This order was fixed to prioritize early collection of high-importance data (drawing task). Within the drawing task, scan order alternated between drawRH and drawLH, starting with the preferred hand. ("Preferred hand" was defined as which hand the participant spontaneously used to write.) This order was fixed because the alternative (starting with the less-dexterous hand) would depress participant morale and increase attrition; with alternating hands to alleviate any potential effects of task order (e.g. it is unlikely that any task order confound would be specific to the 1st, 3rd, and 5th run). All participants who could complete the fMRI drawing task preferred their RH, so all participants received task order: drawRH, drawRH, drawLH, drawRH, drawLH, drawRH, drawLH. Non-MRI data collection times are listed in the "additional data" section. ### Task details Participants drew continuously within presented shapes. "Draw within the lines, and move as quickly as you can, but staying within the lines is more important than speed." The task was presented in a block design, alternating between drawing and rest. Task trials were sorted empirically from easiest to hardest, based on velocity smoothness (of the 45 shape*width possibilities). The 5 middle-difficulty trials were discarded, and the remaining 40 were divided into 10 deciles, each containing 4 shapes. Every task block drew from one decile. Therefore, the 10 drawing blocks in each run can be ranked in order of difficulty, and/or divided into easy blocks (from the easier 20 shapes) vs. hard blocks (from the harder 20 shapes). To allow analysis of these, every run includes 2 alternate _events.tsv files: -task-drawXHeasyhard distinguishes draw-Easy blocks from draw-Hard blocks -task-drawXHdeciles specifies the decile, from 01 (easiest) to 10 (hardest) Participants were familiarized with the drawing task before entering the scanner: they first completed 15 shapes (2-5 minutes) with their preferred hand on an iPad version , followed by 1 run (5.5 minutes) with each hand in a mock scanner. ### Additional data acquired The "phenotype" directory includes following data (and collection times). Items are questionnaires unless otherwise noted. Full item content for all phenotype files is provided in dataDictionary.tsv. ACS-activities.tsv = Activity Card Sort. During visit after scans. ACS-top5.tsv = Activity Card Sort subform on top five activities. During visit after scans. box-and-blocks.tsv = Box and Blocks hand motor assessment. During visit before scans. comfort.tsv = Custom survey on comfort with LH use. Online ≤1 week before scans. DASH.tsv = Disabilities of the Arm, Shoulder, and Hand. Online ≤1 week before scans. data-collection-notes.tsv = General experimenter notes on data collection. demographics.tsv = Demographic information. Online ≤1 week before scans. edinburgh.tsv = Self-reported hand preference (current for all participants; also pre-injury for patients). Online ≤1 week before scans. GAD7.tsv = Generalized Anxiety Disorder-7. Online ≤1 week before scans. MAL.tsv = Motor Activity Log structured interview; amount scale only (patients only). pain-sleep.tsv = Custom survey on pain & sleep. During visit before scans. PHQ9.tsv = Patient Health Questionnaire-9 depression. Online ≤1 week before scans. pni-interview.tsv = Custom structured interview on peripheral nerve injury & patient status (patients only). During visit before scans. pni-symptoms.tsv = Custom questionnaire on symptoms in affected hand in last week (patients only). Online ≤1 week before scans. PROMIS29.tsv = Patient Reported Outcomes-29 questionnaire on health-related quality of life. Online ≤1 week before scans. In addition, each subject's "beh" directory includes the following data, which are detailed in the study directory files "task-X_beh.json": drawLH: high-frequency position data collected from LH precision drawing during fMRI drawRH: high-frequency position data collected from RH precision drawing during fMRI blockBuilding: Block Building Task (BBT) data collected during the visit, prior to fMRI. ### Experimental location Washington University in St. Louis School of Medicine, East Building Bay 2 Prisma 3T. ### Missing data * Three patients were unable to perform the drawing task with their impaired RH, and thus have only LH data (subs 1002, 1019, 1045). * Due to the lengthy scan time (total time in magnet 2+ hours), participants were free to end their participation at any time if they became uncomfortable or tired. As a result, 6 participants are missing some or all restingstate runs, and not all participants have T2 and/or DWI scans. * Block Building Task data was not generated for 4 participants due to personnel limitations (subs 2064, 2066, 2067, 2068), and lost for 1 participant (2029). Some early participants completed only 2 BBT runs; later the protocol was expanded to 4 runs. * Gaps in the subject number sequence do not reflect missing data. Subject numbers were assigned during recruitment, and therefore some were assigned to participants who did not consent or withdrew before data collection.