Computed tomography ventricular size has no predictive value in diagnosing pseudotumor cerebri.

We compared linear indices of computed tomography ventricular size in patients with pseudotumor cerebri and in age-matched controls, and found no difference between these populations. Two physicians showed poor agreement and sensitivity when blindly judging whether patients and controls had "slit-like" ventricles. We conclude that ventricular size on routine CT has no predictive value when evaluating patients with suspected pseudotumor cerebri.