Introduction: Locked-in syndrome (LIS) in children is a rare neurological condition characterised by near-total paralysis despite preserved awareness and cognition. Affected children are unable to interact meaningfully with their environment, leading to loss of autonomy and reduced engagement. Current brain computer interface (BCI) systems for LIS are largely designed for adults. Paediatric-friendly BCI solutions remain scarce. This study presents a proof-of-concept noninvasive BCI system tailored for a paediatric patient. Objectives: To develop and validate a proof-of-concept, training-free EEG paradigm-based BCI using a reduced electrode configuration for a child with LIS. Methods: A non-invasive system was implemented using the OpenBCI Cyton board for EEG signal acquisition and transmission. A five-electrode setup using custom active dry comb electrodes was positioned according to the international 10- 20 system at O1, Oz and O2, TP9 and TP10. The selected paradigm was Steady-State Visually Evoked Potentials (SSVEP). Visual stimuli were presented between 7 Hz and 15 Hz to evoke frequency-specific cortical responses. An adaptive signal processing pipeline was implemented incorporating temporary local multivariate synchronisation index and filter bank canonical correlation analysis for frequency detection. The system output was determined based on frequency-specific correlation responses. Results: The developed signal processing pipeline was evaluated using a publicly available benchmark SSVEP dataset. The system achieved a classification accuracy of 88%. Preliminary experimental trials demonstrated reliable command detection under real-world conditions. Conclusions: We demonstrate the feasibility of a proof-of-concept, training-free EEG paradigm-based BCI using a reducedelectrode configuration for a child with LIS laying the groundwork for increased autonomy .