Magnetic resonance imaging is a promising technological advance used for research and diagnosis of disease. The procedure has no risks, except when uncooperative patients require sedation. Four normal children participated in simulated scans to study the effects of (a) antecedent changes in the imaging environment and (b) operant conditioning of movement inhibition. Changing the environment can decrease movement, but operant contingencies were necessary to decrease movement to a level that, in most cases, would allow the procedure to occur without sedation.DESCRIPTORS: behavioral pediatrics, feedback, magnetic resonance imaging, movement control, patient compliance Magnetic resonance imaging (MRI) is a noninvasive tool for viewing internal anatomy. Patients lie in an endosed, noisy machine for 20 min or more. Movement decreases image resolution, but the exact movement threshold for a failed scan is unknown. The usability of the images can be evaluated only by viewing them after staff and scanner time has been invested. Sedation is ordered in half of all pediatric cases and, in most children under 7 years old, to prevent movement. Sedation adds risk of transient hypoxia, prolonged profound sedation, and respiratory depression (Hubbard, Markowitz, Kimmel, Kroger, & Bartko, 1992).Slifer, Cataldo, Cataldo, and Burke (1989) reported using contingent videotape viewing plus token reinforcement with a 94% success rate for patients meeting a movement criterion of 2 mm per second or less during a 20-min simulated scan. These individuals subsequently cooperated with actual MRI scans without sedation, and the resulting images were judged by medical staff to be of good quality. These subjects were not acutely ill, in pain, or in urgent need of MRI, but induded a range of ages (5 to 46 years), developmental disabilities, and psychiatric diagnoses. However, functional control was not demonstrated. The present study demonstrates functional control over normal children's movement during simulated MRI scans and conducts a partial component analysis of the Slifer et al. (1989) procedure.METHOD: Subjects were normal children who, with parental consent, volunteered for simulated scans. There were 2 boys and 2 girls (5 and 6 years old) from middle-income families with no developmental or medical disorders. The simulated MRI (using an inactive scanner) lasted 20 min (the length of a brain scan). A helmet-like "head coil" covered the child's face, with vision possible through a mirror (standard on MRI scanners). Head supports and safety straps used during MRI were used during simulation. The examination table moved mechanically, and taped scanner noises were played at high volume.Movement was measured with precision potentiometers (Technology Instruments Corp.). Potentiometers have often been used to measure movement in biomedical research (e.g., Melvill-Jones, Guitton, & Berthoz, 1988). The apparatus was connected to the child's forehead by nylon string and an adhesive patch. Head movement activated the potentiometers, which translate...