EEG/(f)MRI measurements at 7 Tesla using a new EEG cap (“InkCap”) Christos E. Vasios, a, ⁎ Leonardo M. Angelone, b Patrick L. Purdon, a Jyrki Ahveninen, a John W. Belliveau, a and Giorgio Bonmassar a a Harvard Medical School, Massachusetts General Hospital, Athinoula A. Martinos Center for Biomedical Imaging, Building 149, 13th Street, Charlestown 02129, MA, USA b Biomedical Engineering Department, Tufts University, Medford, 02155, MA, USA Received 23 December 2005; revised 23 July 2006; accepted 25 July 2006 Available online 10 October 2006 We aimed at improving the signal-to-noise ratio (SNR) of electro- encephalography (EEG) during magnetic resonance imaging (MRI) by introducing a new EEG cap (“InkCap”) based on conductive ink technology. The InkCap was tested with temperature measurements on an electrically conductive phantom head and during structural and functional MRI (fMRI) recordings in 11 healthy human volunteers at 7 T. Combined EEG/fMRI measurements were conducted to study the interaction between the two modalities. The EEG recordings with the InkCap demonstrated up to a five-fold average decrease in signal variance during echo-planar imaging, with respect to a cap made of standard carbon fiber leads. During concurrent EEG/fMRI measure- ments in human volunteers, alpha oscillations were clearly detected at 7 T. Minimal artifacts were present in the T2* and high-resolution structural MR images of the brain parenchyma. Our results show that the InkCap technology considerably improves the quality of both EEG and (f)MRI during concurrent measurements even at 7 T. © 2006 Elsevier Inc. All rights reserved. Keywords: EEG; MRI; Conductive ink; Eddy currents; SAR; Motion sensors Introduction Electroencephalography (EEG) and functional magnetic reso- nance imaging (fMRI) are well-established methods for cognitive and clinical neuroscience. However, each of these methods necessarily has advantages and disadvantages. EEG has a high temporal resolution, but the localization of cerebral sources of signals is complicated by the well-known biophysical challenges (Bai and He, 2005). fMRI is in turn compromised by a poor temporal resolution and the fact that hemodynamic signals are epiphenomena of underlying neuronal activity (Buxton, 2002). Analysis techniques combining imaging data from separate EEG/magnetoencephalo- graphy and fMRI sessions have been developed (Dale et al., 2000), but these approaches suffer from the potential errors caused by intra- subject variability across sessions. However, simultaneous EEG and fMRI recordings offer the possibility of studying brain function with greater detail by combining the high temporal resolution of EEG with the high spatial resolution of the fMRI in images acquired in a single measurement session (Bonmassar et al., 2001, Liu et al., 1998, Vanni et al., 2004). In the last few years, several groups have performed simultaneous EEG and fMRI recordings at 1.5, 2 and 3 T (Allen et al., 2000, Benar et al., 2003, Bonmassar et al., 2001, Comi et al., 2005, Goldman et al., 2000, Iannetti et al., 2005, Ives et al., 1995, Kobayashi et al., 2005, Lemieux et al., 1997, Liebenthal et al., 2003, Matsuda et al., 2002, Mirsattari et al., 2004, Mulert et al., 2005, Scarff et al., 2004) However, the technical challenges of measuring multi-electrode EEG data in the highest magnetic B 0 fields (7 T) have not yet been fully addressed. The interactions of EEG and MRI measurements mutually tend to decrease the signal-to-noise ratio (SNR) of the data of each modality. EEG leads can interfere with the radiofrequency (RF) field by detuning the coil used in MR imaging, resulting in a global attenuation the RF signal received (Schomer et al., 2000). fMRI imaging can, in turn, interfere with EEG recordings because of (a) echo-planar imaging (EPI) noise and (b) ballistocardiogram noise resulting from lead motion by heartbeat, breathing, and other vibrations. The EPI noise refers to noise currents in the EEG leads and in the human head induced both by RF imaging pulses and by pulsed gradient magnetic fields (Huang-Hellinger et al., 1995). The ballistocardiogram noise, in turn, which stems from the altered loop surface area by motion, induces currents that are the primary component of noise in the EEG signal inside the static field (Allen et al., 2000, Hill et al., 1995). This artifact is especially pronounced in the very high magnetic B 0 fields (e.g., 7 T) that are used in high- resolution MRI. The potential risk for subjects participating in combined EEG and fMRI measurements is an even more essential complicating factor. The RF fields can also induce local currents potentially resulting in tissue heating (Chou et al., 1996, Lemieux et al., 1997). For this www.elsevier.com/locate/ynimg NeuroImage 33 (2006) 1082 – 1092 ⁎ Corresponding author. E-mail address: cvasios@nmr.mgh.harvard.edu (C.E. Vasios). Available online on ScienceDirect (www.sciencedirect.com). 1053-8119/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2006.07.038