Anatomical variability of the lateral frontal lobe surface: implication for intersubject variability in language neuroimaging Heike Juch, a Ivan Zimine, b, * Mohamed L. Seghier, b Franc ¸ois Lazeyras, b and Jean H.D. Fasel a a Department of Morphology, Division of Anatomy, Medical School, University of Geneva, 1, Geneva, CH-1211, Switzerland b Department of Radiology, Geneva University Hospitals, Geneva, CH-1211, Switzerland Received 12 May 2004; revised 13 August 2004; accepted 25 August 2004 The lateral surface of the frontal lobe shows functional activation in a large number of language related tasks. Group analyses, however, demonstrate remarkable intersubject variability of activation. There are different sources for functional variability, anatomical variability being considered as one of them. The aim of the present study therefore was to qualitatively and quantitatively investigate the anatomical variability of the lateral frontal lobe surface and to search for reliable and stable landmarks connected to language functions. MRIs of 23 healthy right-handed subjects were investigated using the publicly available software bAnatomist/BrainVISAQ. After standardiza- tion of the brains (SPM) and sulci identification, the most frequent pattern was determined and the variance of selected landmarks calculated. The variability of the lateral frontal lobe surface is remarkable, particularly in the prefrontal region. Relatively stable landmarks were selected as follows: (1) connection between the superior frontal sulcus (SFS) and the superior precentral sulcus (SPCS); (2) connection between the inferior frontal sulcus (IFS) and the inferior precentral sulcus (IPCS); (3) inferior end of the precentral sulcus (PCS); and (4) origin of the ascending ramus (AscR) of the Sylvian fissure (SYF). The variability (standard deviation) of the spatial coordinates along the 3 axis of these landmarks after normalization ranged from 2.5 to 5.7 mm. The present study demonstrates that intersubject variability of selected landmarks of the frontal lobe surface remains notable even after spatial normalization of the brains. These results support the concept that anatomical variability is a relevant source of functional variability. We therefore suggest to express functional activation in relation to land- marks obtained from individual anatomy. This approach may contrib- ute to a better analysis of the differences between individuals. D 2004 Elsevier Inc. All rights reserved. Keywords: Anatomical variability; Cerebral sulci; Frontal lobe; Functional variability; Language; Broca’s area Introduction Language-related functions were among the first to be ascribed a specific location in the human brain (Broca, 1861). The bclassical modelQ of language function, based on observations of aphasic patients, proposes an area in the left inferior frontal gyrus (IFG) for planning and executing speech and writing, named after Broca (1861). Functional magnetic resonance imaging (fMRI) has confirmed and enlarged the knowledge about the role of the frontal cortex in language processing. Broca’s area is reproducibly activated in fMRI (Binder et al., 1997; Bookheimer, 2002; Brannen et al., 2001; Herholz et al., 1996; Hund-Georgiadis et al., 2002; Nolde et al., 1998; Poldrack et al., 1999; Rutten et al., 1999; Stippich et al., 2003). However, fMRI demonstrated the involvement of numerous other cerebral areas, in particular in the premotor and prefrontal cortex. These areas are activated by well defined and validated tasks encompassing phonological, semantic, syntactic, orthographic, and lexical components (Billingsley et al., 2001; Friederici et al., 2000; Frost et al., 1999; Fujimaki et al., 1999; Hund-Georgiadis et al., 2002; Kuperberg et al., 2000; Poldrack et al., 1999; Seghier et al., 2004; Stippich et al., 2003). As far as hemispheric dominance of language processing is concerned, the reliability of fMRI to determine lateralization was shown in several investigations (Carpentier et al., 2001; Frost et al., 1999; Hund-Georgiadis et al., 2002; Rutten et al., 2002a; Seghier et al., 2004). This fMRI assessing of language lateralization corroborates well with the invasive intracarotid amobarbital procedure (Wada test, Wada and Rasmussen, 1960; Binder et al., 1996; Fernandez et al., 2003; Lehericy et al., 2000; Rutten et al., 2002b; Woermann et al., 2003). However, functional imaging studies have generated a large amount of sometimes inconsistent results (Cabeza and Nyberg, 2000). With the same task, results in terms of frequency of occurrence, localization, intensity, and extent of activation can largely differ from one study to another. Reasons for this inconsistency are presumed to be (1) the subject factor (handness, gender, age, native language), (2) the paradigm itself (production or comprehension tasks, language component, presentation modality), 1053-8119/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2004.08.037 * Corresponding author. Department of Radiology, Geneva University Hospitals, 24, rue Micheli-du-Crest, Geneva, CH-1211, Switzerland. E-mail address: ivan.zimine@sim.hcuge.ch (I. Zimine). Available online on ScienceDirect (www.sciencedirect.com). www.elsevier.com/locate/ynimg NeuroImage 24 (2005) 504 – 514