Trials of mirror therapy using mirror visual feedback (MVF) require appropriate sham controls, but visually similar control conditions can nonetheless elicit mirror‑related cortical responses. This protocol outlines an acute mechanistic screening study using functional near‑infrared spectroscopy (fNIRS) to identify which prespecified sham material most closely approximates a neural "neutral" signature compared with true MVF during a single exposure. The aim is to reduce the risk of active sham effects in future clinical trials by empirically selecting the least mirror‑engaging control material.
This is a single‑centre, within‑participant, randomised crossover study conducted at Wuhan Wuchang Hospital (Wuhan, China). The target sample is healthy adults aged 18–35 years. Each participant completes four experimental conditions in a single visit, with condition order counterbalanced using a Latin‑square schedule to control for order effects and reduce between‑participant variability. The four conditions comprise one true MVF condition (C1) and three prespecified sham materials (C2–C4).
Participants perform a standardised grasping task while wearing the fNIRS optode montage. The four conditions are administered in a single session, each presenting a different visual material or arrangement intended to replicate true MVF (C1) or represent alternative sham materials (C2–C4). Task timing and instructions are standardised across conditions. Blinding of participants to condition identity is attempted, with a blinding assessment included to document any perceptual differences or expectancy effects. Exact material definitions and the frozen channel‑to‑ROI mapping are recorded in the Supplementary materials to allow operational reproducibility.
fNIRS data are acquired throughout the grasping task. Online acquisition‑time quality control uses signal‑quality indices (SCI) and coefficient of variance (CV) thresholds with prespecified re‑acquisition rules to address poor recordings during the session. Preprocessing steps and artifact handling are prespecified in the protocol; despite these measures, the protocol acknowledges that fNIRS remains susceptible to motion artefact, scalp‑coupling variability, and physiological noise, which may reduce sensitivity.
The primary neurophysiological outcome is region‑of‑interest task‑evoked change in oxygenated haemoglobin (ΔHbO) within prespecified ROIs: premotor cortex (PMC) and primary sensorimotor cortex (SM1/M1). Task‑evoked responses will be estimated primarily using general linear model (GLM)–derived beta estimates. Condition effects will be assessed with linear mixed‑effects models including prespecified contrasts. Holm multiplicity adjustment will be applied to control family‑wise error across contrasts. A prespecified neutrality decision rule will be used to rank the three sham conditions by their neural similarity to a neutral signature relative to true MVF.
The protocol includes several operational safeguards to improve reproducibility: standardised task timing and instructions, frozen channel‑to‑ROI mapping, predefined SCI and CV thresholds with re‑acquisition rules, and detailed records of sham material definitions. Key supplementary materials documenting these items will be provided with the manuscript to support transparent replication.
Ethics approval was obtained from the Ethics Committee of Wuchang Hospital Affiliated to Wuhan University of Science and Technology (Approval No.: 2025-112-01). The trial is registered in the Chinese Clinical Trial Registry (ChiCTR2600116634). The authors state that de‑identified participant data will be made available in accordance with ethics approval and institutional policy via a public repository after publication of the primary results; access will require a methodologically sound proposal and a data use agreement prohibiting re‑identification.
Strengths of the study design include the within‑participant randomised crossover structure, which reduces between‑participant variability and is efficient for acute mechanistic screening of sham conditions. Prespecifying sham conditions, ROIs, contrasts, a neutrality decision rule, and Holm multiplicity control reduces analytic flexibility and supports transparent interpretation. Operational reproducibility measures are detailed to facilitate replication.
Limitations noted by the authors include the single‑centre, healthy‑participant, single‑session paradigm, which may limit generalisability to clinical stroke populations and to longer‑term therapeutic effects. Blinding may be imperfect because perceptual differences between materials can affect expectancy and attention; although blinding assessment is included, residual bias is possible. Finally, despite prespecified QC and preprocessing, residual fNIRS artefacts from motion, scalp coupling, or physiological signals may remain and reduce sensitivity.
The protocol and subsequent results manuscript will be submitted for publication, with key supplementary materials provided as online appendices or supplements as required by the target journal. The authors declared no competing interests. The study is funded by the Wuhan Natural Science Foundation – Key Clinical Research Program for Municipal Medical Institutions (2026020301040188).