A New Brain-Shift Model for Neurosurgery with Fronto-Temporal Craniotomy

Neuronavigation systems have become standard neurosurgical tool; however, there is a major concern to be solved. Brain shift occurs during surgery, which compromises the system’s accurate anatomical representation. We developed and evaluated a convenient new method for a rapid intraoperative correction of brain shift during neuronavigation. We assessed four patients (4 females; mean age 61.0 years) that underwent a fronto-temporal craniotomy. Each voxel movements of CT images in the brain parenchyma are analyzed as displacement vector using neuronavigation system, and a new free-form deformation method was established among these patients and verified in another one. The shape concordance rate between the actual intraoperative CT image and CT image, which was corrected by our model, was 75.1%. On the basis of nonlinear geometric algorithms that involve intraoperative measurements of anatomical landmark positions, our model might beuseful especially in pterional craniotomy, which is one of the most common approaches employed in neurosurgery. Future improvements and further accumulation of patients’ data will enable our model to be applied to a variety of surgeries.