Abstract
For nearly 90 years, notions of the brain have been inextricably associated with a homunculus that has become embedded within medical education as the precise representation of rolandic cortical function. We sought to define the history, evolution, accuracy, and impact of this pictorial means of showing cortical representation. We mathematically defined the evolutionary accuracy of appropriate homunculi using image analysis techniques for all points defined by Penfield, Boldrey, Rasmussen, Jasper, and Erickson, calculating perpendicular distances and defining areas and distributions of rolandic and sylvian regions labeled for sensory and motor activity with comparison with all homunculi. Prerolandic sensory representation composed 13%-47% of total sensory area (mean, 29%); postrolandic motor representation composed 15%-65% of total motor area (mean, 31%). Discrepancy between cortical perpendicular length attributed to a particular function on 1937 diagrams was greater than that attributed on the 1950 homunculus (motor: mean, 74%; range, 63%-96%; sensory: mean, 66%; range, 17%-92%) (P < 0.05). The homunculus, if truly drawn according to cortical mapping evidence, could never have been recognized as near humanoid, yet it has attained epic educational and practical longevity.