Dystonia in cerebral palsy is debilitating but underdiagnosed precluding targeted treatment that is most effective if instituted early. Deep gray matter injury is associated with dystonic cerebral palsy but is difficult to quantify. Objective and clinically feasible identification of injury preceding dystonia could help determine the children at the highest risk for developing dystonia and thus facilitate early dystonia detection.

We examined brain magnetic resonance images from four- to five-day-old neonates after therapeutic hypothermia for hypoxic-ischemic encephalopathy at a tertiary care center. Apparent diffusion coefficient values in the striatum and thalamus were determined using a web-based viewer integrated with the electronic medical record (IBM iConnect Access). The notes of specialists in neonatal neurology, pediatric movement disorders, and pediatric cerebral palsy (physicians most familiar with motor phenotyping after neonatal brain injury) were screened for all subjects through age of five years for motor phenotype documentation.

Striatal and thalamic apparent diffusion coefficient values significantly predicted dystonia with receiver operator characteristic areas under the curve of 0.862 (P = 0.0004) and 0.838 (P = 0.001), respectively (n = 50 subjects). Striatal apparent diffusion coefficient values less than 1.014 × 10⁻³ mm²/s provided 100% specificity and 70% sensitivity for dystonia. Thalamic apparent diffusion coefficient values less than 0.973 × 10⁻³ mm²/s provided 100% specificity and 80% sensitivity for dystonia.

Lower striatal and thalamic apparent diffusion coefficient values predicted dystonia in four- to five-day-old neonates who underwent therapeutic hypothermia for hypoxic ischemic encephalopathy. Objective and clinically feasible neonatal brain imaging assessment could help increase vigilance for dystonia in cerebral palsy.