Asian Journal of Medicine and Health

Magnetic resonance imaging (MRI) is central to paediatric diagnostic radiology because it avoids ionising radiation, yet its dependence on a motionless subject over several minutes to tens of minutes is poorly matched to the developmental capacities of infants and young children. Historically, this mismatch has been resolved through pharmacological sedation or general anaesthesia, practices that carry documented risks of cardiorespiratory adverse events, incur substantial cost, and have become the subject of regulatory concern following signals of anaesthetic neurotoxicity in the developing brain. Over the past two decades, two largely parallel bodies of work have emerged to reduce reliance on pharmacological immobilisation: patient-centred behavioural and environmental preparation strategies, including child life specialist coaching, mock scanner rehearsal, and parental involvement; and technological advances in acquisition and reconstruction, including accelerated protocols, deep learning image reconstruction, motion-robust sequences, and acoustically quiet scanning. This critical narrative review synthesises the accessible peer-reviewed literature on both domains, evaluates the strength and limitations of the underlying evidence, and examines how these strategies interact within contemporary care pathways. The review finds convergent evidence that structured behavioural preparation and the feed-and-wrap technique in neonates substantially reduce sedation utilisation without materially compromising diagnostic image quality, while acknowledging that most supporting studies are single-centre, non-randomised, and vulnerable to selection bias favouring more cooperative children. Evidence for motion-robust acquisition and deep learning reconstruction is more recent and dominated by retrospective feasibility work rather than outcome-linked trials. The relationship between brief anaesthetic exposure and long-term neurodevelopment remains reassuring in the only available randomised trial, while cohort associations involving repeated or prolonged exposure remain confounded by underlying illness. The review argues that no single intervention resolves the motion problem in isolation; an integrated, age- and indication-stratified pathway combining preparation, environmental design, and motion-robust technology offers the most defensible route to reducing unnecessary sedation while preserving diagnostic yield. Priorities for future research include multicentre randomised comparisons of preparation modalities, prospective diagnostic-outcome validation of deep learning reconstruction, and economic evaluation outside single-institution settings.