Frameless stereotaxy in the nonhuman primate

With the advent of magnetic resonance imaging (MRI), it is possible to obtain high-resolution anatomical images of the monkey brain. Accuracy, however, is lost in the laboratory or surgical setting when the localization of brain structures depends on nonstereotaxic tracking methods. Here we present an image-guided stereotaxic system that is able to localize and access anatomical brain structures using the monkey's MRI. This system, which is also known as frameless stereotaxy, is capable of computing the relation of the physical “real space” of the monkey's head to the corresponding image space, while a position sensor enables the tracking of the animal's head and the localization of brain areas and favorable paths to targets within the brain using real time display software. Surgical procedures make use of an adjustable upright chair and a surgical headclamp instead of the traditional restrictive head holder with ear bars. This novel system allows for the flexible positioning of the animal and the ability to reach areas of the brain that were difficult to access in the past. The headclamp also serves as a tool holder, which in the present application guided a cannula of retrograde tracer to the desired location in the frontal lobe. Histological examination of the brain showed that the injection reached the target site, and tests using an MRI compatible phantom demonstrated that the precision of the system in bringing an injection to target is less than 1.2 mm. This system can be used to inject accurately tracers for anatomical tract-tracing, to make precise lesions, and to position electrodes for electrophysiological studies.

Introduction

Magnetic resonance imaging (MRI) has greatly improved our ability to view the brain and its component structures in a noninvasive manner (for review, see Brownell et al., 1982, Haacke et al., 1999, Moonen et al., 1990, Westbrook and Kaut, 1998). Although MRI scans are routinely performed on nonhuman primates, using these scans to perform precise experimentation is limited due to the lack of real time feedback to the researcher about anatomical structures encountered during an invasive procedure in the laboratory or surgical suite. With the enormous cost in time and money in acquiring and training monkeys, and the need to reduce all unnecessary risks to animals, measures should be taken to ensure that the techniques used in monkey research are performed as accurately as possible.

Traditionally, a stereotaxic frame has been used to reach desired anatomical targets in the monkey brain, based on existing generalized stereotaxic atlases (Paxinos et al., 2000, Snider and Lee, 1961, Szabo and Cowan, 1984) or MRI images using a coordinate system for three axes (x, y, z) (Alvarez-Royo et al., 1991, Maciunas and Galloway, 1989, Rebert et al., 1991, Sapolsky et al., 1990). Making use of frame-based stereotaxic coordinates to localize brain structures with MRI originated in surgery for humans (Lunsford et al., 1986, Peters et al., 1986), its main advantage being accuracy when image distortion is eliminated (Dormont et al., 1994). Although a benefit to the frame-based method is that it provides structural support for tools that are needed for the different surgical procedures, the frame also limits access to large parts of the brain. Another significant limitation of a frame-based stereotaxic system in both human (Olivier and Bertrand, 1983, Popovic and Kelly, 1993, Ross et al., 1996, St-Jean et al., 1998) and monkey (Alvarez-Royo et al., 1991, Maciunas and Galloway, 1989, Saunders et al., 1990) is that the subject must be scanned with the actual frame in place. In the case of nonhuman primates, especially larger animals, this can be cumbersome as the animal must be placed in an unnatural position in the scanner and within the frame to obtain usable images. Although several researchers have used glass-filled copper sulfate beads embedded under the scalp to calculate stereotaxic coordinates without using a frame during MRI scanning, a standard stereotaxic frame is still necessary for all subsequent surgical procedures (Rebert et al., 1991, Sapolsky et al., 1990). It is also necessary with a frame-based technique to determine target approach and record coordinates for each target at a presurgical planning stage. Once this has been done, it cannot be changed without recalculating target approach. In contrast, with the frameless stereotaxic system presented here, the animal is simply scanned with several relocatable reference points (fiducials) in the standard supine position. When the fiducials are chronically implanted, they can be used to coregister the animal's head in reference to a 3D position sensor, allowing the researcher to proceed with any procedure or manipulation at a later date following the MRI scan. An advantage of the frameless stereotaxic system is that the registered coordinates and images are stored on file and only a few minutes are needed to calibrate the animal to its own previously acquired MRI scan. It is then possible to localize the target region and determine target approach just before surgery. The frameless technique is flexible enough that if necessary, the target approach can easily be changed to accommodate a different entry point during the surgical procedure. The frameless technique also allows the head of the animal to be positioned in virtually any position while still allowing access to the desired target area.

This report presents a novel approach using a frameless stereotaxic system to enable the precise localization of a target area in the monkey brain and to guide accurately a cannula to inject an anatomical tracer to the desired location. The accuracy of this technique was supported with histological evidence, and the overall system error was investigated by measuring target registration error with a custom-made MR imaging phantom.