Experimental Radiofrequency Brain Lesions: A Volumetric Study

OBJECTIVE This study describes the production, under strictly standardized and controlled conditions, of radiofrequency lesions with identical neurogenerator settings: in vitro in two different albumin solutions (nongelatinous and gelatinous) and in vivo in the thalamus of the pig. METHODS The radiofrequency lesions were investigated in vitro by the use of a specially designed video system and in vivo by magnetic resonance imaging. Moreover, the size of the in vivo lesions was estimated with the use of histological sectioning. The statistical analysis included the calculation of a correlation coefficient for the length, width, and volume for each lesion estimation. RESULTS A high correlation (R = 0.96, P < 0.005; n = 14) was found between clot sizes in the two albumin solutions. Albumin clots generated in gelatinous albumin showed systematically larger volumes. In the pig, two concentric zones were seen in all magnetic resonance images and all histological preparations. The width correlation of the completely coagulated brain tissue (inner zones) was R = 0.94, P < 0.005, and n = 7. The corresponding correlation between magnetic resonance images and gelatinous albumin was R = 0.93, P < 0.005, and n = 7. As a rule, the in vitro clots were smaller than the outer zone but larger than the inner zone of the magnetic resonance imaging-recorded lesions for all of the electrode and temperature combinations tested. In vivo lesions generated with the same electrode and parameter settings showed high reproducibility. CONCLUSION The value of presurgical electrode tests to validate the electrode function and lesion size in vitro has become evident in this study, which shows a high correlation between the in vitro albumin clots and the in vivo lesions observed on magnetic resonance images.