Finding studies
Finding studies
Take this into the appointment.
Saves the questions and what to expect into your notes, next to the visit they belong to.
Manuel J Marte, PhD, CCC-SLP
CONTACT
Melissa D Stockbridge, PhD, CCC-SLP
CONTACT
Lead
Johns Hopkins University
With
About one third of people with aphasia after stroke respond poorly to language therapy, and current clinical tools cannot identify these individuals early, when recovery is steepest. This study pairs a recovery prediction model with a targeted treatment. Aim 1: A machine learning model trained on acute-phase data from the parent longitudinal study (Johns Hopkins IRB NA\_00042097; neuroimaging, cognitive-linguistic testing, demographics) predicts each participant's 12-month Western Aphasia Battery-Revised Aphasia Quotient (WAB-R AQ). Participants predicted to remain below 93.8 (aphasia not expected to resolve) are eligible. Prediction accuracy is evaluated prospectively against observed 12-month scores (mean absolute error, correlation, and sensitivity, specificity, positive predictive value and F1 for the below-93.8 classification). Aim 2: Eligible participants are randomized 1:1 to early intervention (treatment beginning about 3 months post-stroke) or wait-list control (usual care until about 6 months post-stroke, then the same treatment). Treatment is 15 sessions, up to 5 per week over about 3 to 5 weeks. In each session, cathodal tDCS (2 mA, 25 minutes; cathode over the right cerebellum, 1 cm below the inion and 4 cm lateral; anode over the right deltoid; Soterix Medical 1×1 device) is delivered during the first 25 minutes of 45 to 60 minutes of computer-delivered language therapy targeting picture naming, auditory comprehension, reading and spelling, with a speech-language pathologist present throughout. Sessions may take place at a Johns Hopkins site or at home with a study team member present. The primary comparison is change in WAB-R AQ from baseline (about 3 months post-stroke) to 6 months post-stroke between the treated early group and the untreated wait-list group. Secondary measures include naming of trained items, discourse (content units and syllables per content unit in picture description), functional communication (Communication Activities of Daily Living, Third Edition), emotional well-being (General Health Questionnaire), quality of life (Stroke and Aphasia Quality of Life Scale), and resting-state functional connectivity measured with functional near-infrared spectroscopy (fNIRS). Participants are followed to 12 months post-stroke. As a pilot, the study is designed to show feasibility of the predict-then-intervene approach and to estimate effect sizes for a future fully powered, sham-controlled trial.
Age
18–any
Sex
ALL
Healthy volunteers
Not accepted
