Finding F3: A comparative analysis of scalp-based methods for TMS coil positioning

Background: The left dorsolateral prefrontal cortex (DLPFC) is the most common target for transcranial magnetic stimulation (TMS) treatment of depression and many other neuropsychiatric disorders. The electrode position F3 (EEG F3) has been shown to be an adequate scalp heuristic for targeting the left DLPFC, and the Beam F3 method has become the standard approach to approximate EEG F3. However, the accuracy of the Beam F3 method has recently been questioned, and the scalp geometry-based parameter (SGP) method has emerged as an alternative. Objective: To compare the performance of the Beam F3 and SGP methods for localization of EEG F3. Methods: We created MRI-based head mesh models of 114 participants derived from the Human Connectome Project. Using surface based measurements, we first defined EEG F3 according to the 10/10 EEG system and then assessed the relative positions of Beam F3 and SGP F3 points. Localization performance was quantified for each participant in terms of distance and angle on the mesh surface. Inter-individual differences were explored related to head size, head shape, gender, age, and clinical phenotype (controls vs. patients). Results: A significant difference was found between the Beam F3 and SGP F3 points, with a mean distance from electrode position F3 of 3.58 mm and 1.23 mm, respectively. Importantly, head shape had a significant negative correlation with localization performance, but only for the Beam F3 approach, while no correlation was found using the SGP method. Gender, age and phenotype had no effects on localization performance. Conclusions: Our results demonstrate shortcomings of the Beam F3 method and reveal that the SGP approach enables more accurate localization of EEG F3. It is critical to make a distinction between the EEG F3 and the Beam (F3) point to ensure methodological clarity, and to start evaluating the potential effects of this discrepancy on the efficacy of TMS applications in practice.