SQUID-based systems for co-registration of ultra-low field nuclear magnetic resonance images and magnetoencephalography A.N. Matlashov ⇑ , E. Burmistrov, P.E. Magnelind, L. Schultz, A.V. Urbaitis, P.L. Volegov, J. Yoder, M.A. Espy Los Alamos National Laboratory, P.O. Box 1663, MS-D454, Los Alamos, NM 87545, USA article info Article history: Accepted 21 April 2012 Available online 30 April 2012 Keywords: MRI ULF Microtesla Low-field MRI MEG SQUID abstract The ability to perform magnetic resonance imaging (MRI) in ultra-low magnetic fields (ULF) of 100 lT, using superconducting quantum interference device (SQUID) detection, has enabled a new class of mag- netoencephalography (MEG) instrumentation capable of recording both anatomical (via the ULF MRI) and functional (biomagnetic) information about the brain. The combined ULF MRI/MEG instrument allows both structural and functional information to be co-registered to a single coordinate system and acquired in a single device. In this paper we discuss the considerations and challenges required to develop a com- bined ULF MRI/MEG device, including pulse sequence development, magnetic field generation, SQUID operation in an environment of pulsed pre-polarization, and optimization of pick-up coil geometries for MRI in different noise environments. We also discuss the design of a ‘‘hybrid’’ ULF MRI/MEG system under development in our laboratory that uses SQUID pick-up coils separately optimized for MEG and ULF MRI. Published by Elsevier B.V. 1. Introduction Magnetoencephalography (MEG) and magnetic resonance imaging (MRI) were made commercially available approximately at the same time, in the mid 1980s, and have coexisted for more than three decades as two technically incompatible methods. While MEG pushed toward measurements at the femto-Tesla level, MRI was pushing toward operating in magnetic fields some 15 orders of magnitude higher, at the Tesla range. Both methods, however, depended significantly on technical advances in superconductivity. Basically, MEG was born as a consequence of the invention of the SQUID – the superconducting quantum interference device. MRI became a mature and widespread method after the development of high uniformity Tesla-range superconducting magnets. Both methods played crucial roles in research and diagnostics of the brain over the last three decades. However, until recently they were used separately because of the disparity in magnetic field strengths. MRI provided morphological images of the brain and MEG provided functional images related to bioelectrical brain activity. The images could be superimposed by using co-registration techniques. How- ever, practically achieving adequate co-registration has often pro- ven challenging. The idea of making one machine for both MEG and MRI ap- peared impossible until two key ideas crossed paths. The first idea came from a paper published by Macovsky and Connolly in 1993 that pointed out that the field-cycling pre-polarization technique enables the development of inexpensive low-field MRI machines [1]. The second idea came from advances in SQUID instrumenta- tion. A few groups experimented with SQUIDs for the measure- ment of NMR signals at low Larmor frequencies, usually a few kHz. This technique was pioneered [2] and significantly developed by John Clarke and his colleagues at Berkeley. They recorded the first ever MR images of plants and a human forearm and fingers at 132 lT field [3], which corresponds to a 5.6 kHz Larmor fre- quency. Such achievements in SQUID-based MRI gained the inter- est of many MEG groups over the world. The idea of using SQUID arrays for simultaneous detection of both MEG and brain MRI, which could reasonably improve the superposition of images from the two different modalities, and eliminate the requirement for two costly devices, is extremely attractive. The first proof-of-principle results in recording simultaneous MEG and MRI signals were achieved in 2004 [4,5]. In this work, the somatosensory evoked magnetic response from the human brain was recorded truly simultaneously with the free induction decay (FID) signal at 268 Hz. During such experiments it became clear that simultaneous recording of such signals is very difficult to perform because of the large low frequency noise arising from the external magnetic field needed for the NMR precession. This noise was caused by micro-vibrations of the SQUID gradiometer in the NMR field and it seriously distorted the MEG signal. In later experiments MEG and MRI signals were recorded sequentially and the external field and gradients were zeroed during the MEG recording. 0921-4534/$ - see front matter Published by Elsevier B.V. http://dx.doi.org/10.1016/j.physc.2012.04.028 ⇑ Corresponding author. Tel.: +1 505 665 6183; fax: +1 505 665 2549. E-mail address: matlach@lanl.gov (A.N. Matlashov). Physica C 482 (2012) 19–26 Contents lists available at SciVerse ScienceDirect Physica C journal homepage: www.elsevier.com/locate/physc