VanderbiltUniversitySchool of Engineering
Department of Biomedical Engineering
Senior Design Final Report
3T NeuroImaging Primate Chair
Jennifer R. Pryweller
April 25, 2006
Advisor:
Professor Malcolm J. Avison
Vanderbilt University Institute of Imaging Science
Abstract
The synergistic activity of large neural populations remains largely unexplored in the visual cortex. Studies explore the brains of rhesus macaques by presenting the monkey with a visual stimulus while simultaneously collecting imaging and electrophysiological data from the occipital lobe.A primate chair will be constructed to restrain awake rhesus macaques during BOLD fMRI brain scans. Based on MR compatibility and other specific factors, the chair will be constructed using polycarbonate, ABS, brass and nylon. The primate chair, while conforming to the highest of ethical standards, provides the macaque with physical comfort and minimal emotional distress. It also promotes a high level of safety for both the monkey and the researcher. Most importantly, the chair is easily modifiable, allowing for the future addition of functional components.
Introduction
The synergistic activity of large neural populations remains largely unexplored in the visual cortex. Hemodynamic metabolic approaches to exploring this activity are based on the fact that energy metabolism (measured as a function of cerebral blood flow and volume change) is coupled to neuronal activity. Studies explore the brains of rhesus macaques by presenting the monkey with a visual stimulus while simultaneously collecting imaging and electrophysiological data from the occipital lobe. Specifically, the correlation of BOLD (blood oxygen level dependent) fMRI (functional magnetic resonance imaging) brain data and neuron action potential signal localization is used to map the visual cortex. Since these investigations require the use of awake monkeys, it is necessary to develop a device to restrain animal movement in an ethical manner while performing simultaneous data acquisition.
A primate chair will be constructed to accommodate fMRI brain scans in awake rhesus macaques. The chair will include adjustable features to provide for variable monkey size. The neck plate diameter and angle will also be adjustable to account for different monkey sizes. Key features of the chair include a removable waste tray, a reward and reinforcement system, stereotaxic devices and electrophysiological monitoring systems. A juice reward system will be automated so that upon correct response to visual task stimuli an air pump will allow the monkey to drink a set amount of juice. In addition, a tone will sound that the monkey is trained to recognize as positive reinforcement. Negative audio tone reinforcement will be used for incorrect visual response. Stereotaxic devices that serve to assist the monkey in holding still (eliminating artifacts in the scan and the need for post-processing motion correction) include and a head post and head bar. Electrophysiological measures of interest during an awake monkey scan include heart rate, blood pressure, body temperature and facial activity. Electrodes will be placed above each eyebrow and on the side of each jaw to monitor muscle twitches in the face. A video camera will be put in place to monitor the monkey's facial and visual responses. A custom circuit will be designed to accommodate electrodes placed in the visual cortex of the brain during scans to identify and monitor individual neurons that are activated in response to specific visual tasks. The diagram in Figure 1 was created to outline and organize the functional components of the entire system.The entire chair will be designed with the monkey's comfort as forethought while conforming to high ethical standards. Holes will be drilled in the walls of the chair to allow for air flow. Since the monkey is trained to sit in the sphinx position for the duration of the scan, the chair will be ergonomically designed to accommodate comfort.
Due to the large size of the project a more specific goal was set to be accomplished in conjunction with the senior design project. Based on design constraints and specifications the design and construction of the physical chair, while accounting for the addition of functional components, was the established as the short term goal.
Previous Primate Chair Designs
Previous designs of primate chairs to meet similar study objectives include those of Wim Vanduffel and Mark Pinsk. The design of Vanduffel’s chair (Figure 2) for use in his 1991 BOLD fMRI studieswas not the primary focus of his experiments, and therefore published data describing the monkey chair is very limited. His chair was constructed to fit a 3-4kg rhesus macaque, sitting in the sphinx position. A head post, permanently fixed on the monkey’s head, was use to immobilize the monkey’s head by attaching it to a head bar built into the chair. The head bar was supported by a cross bar and head frame (Figure 3). This technique of head stabilization was adequate, but BOLD fMRI is so sensitive to movement that better signal could be obtained by further stabilizing the monkey’s head and body. Even with the head held completely still, subtle body movements during scanning can still have a large enough effect on the brain’s magnetic field to cause artifacts in the scan image, resulting in inaccurate measurements and conclusions. The materials used for the construction of Vanduffel’s chair are not published. Though contact was made with Vanduffel in which he agreed to share the blueprints of the chair, subsequent efforts to obtain the plans proved fruitless.
Mark Pinsk, of PrincetonUniversity published a paper in 2005 relating methods of fMRI imaging in awake monkeys. His construction of a primate chair was much more detailed in his published work than that of Vanduffel. Pinsk constructed the chair out of all magnetic resonance (MR) compatible materials, from the siding of the chair, to the implanted head post and all the bolts and screws along the way. His chair was constructed from a cast acrylic tube, 62cm in length and 28cm in diameter. The rear plate was constructed of polycarbonate and left a hole to allow for extension of the monkey’s tail. The front end of the tube was cut at an angle of 30 degrees with respect to the direction of the chair’s length. Several plates were constructed to fit at the front of the tube, each with a large hole in the middle, suitable to encompass the circumference of a monkey’s neck. A flat piece of delrin, with holes, was placed across the bottom of the tube to allow for drainage and separate the monkey from its waste. The head immobilization system was made out of delrin and included a head bar, cross bar and head frame. Two support posts were added to further immobilize the head, and were placed perpendicular to the cross bar on either side. Pinsk’s design, while it is a progression from Vanduffel’s, leaves room for improvements to be specified in this design project.
Crist Instrument is a major manufacturer for primate research equipment. Their design of an MRI compatible primate chair is seen in Figure 5a. They also manufacture a commonly implanted head post known as a footed head post. The VUIIS cohort will be implanted with head posts similar to those in Figure 5b, but made of MR compatible material. It is not sufficient to purchase a primate chair from Crist for use in these experiments because the Crist chair is not large enough in diameter, nor would it snuggly fit the bore of the 3T Achieva MRI scanner.
Ethical Considerations
Rhesus macaques were chosen for use in these studies because they are the most anatomically and physiologically similar to humans. They are also easy to maintain and breed in captivity. The Institutional Animal Care and Use Committee (IACUC) at every research institution is charged with reviewing animal welfare issues and approving all research involving the use of animal subjects. IACUC also provides training for all researchers to promote awareness of ethical considerations and make them more accountable for their actions. The Vanderbilt University IACUC approved all portions of the proposed visual studies, including the construction and use of a primate chair to restrain awake monkeys. The student involved in designing and constructing the primate chair voluntarily took an IACUC research training course on the ethical care and use of animals. A certificate was obtained at the end of the course after passing an examination.
A group of physicians, statisticians, researchers, community advocates and others come together to form the Institutional Review Board (IRB) at all research institutions. The IRB reviews studies proposed by investigators and ensures the ethical treatment and rights of all subjects. Just as it was approved by IACUC, all portions of these proposed studies have been approved by the IRB at VanderbiltUniversity.
Methods
Designs by Pinsk and Vanduffel, while both adequate for some applications, were not suitable for the needs of the Vanderbilt University Institute of Imaging Science in their desire to restrain awake rhesus macaques for visual studies. Unique constraints were imposed upon the design of a primate chair for the VUIIS by the physical dimensions of the bore of the scanner, the specification of the available rhesus cohort, and the inherent constraints of magnetic compatibility in an MRI scanner.
Several constraints were imposed by the bore of the magnet since every MRI scanner is slightly different in shape and size. A Philips Achieva 3T MRI scanner is seen in Figure 6, with the bore (inner radius) shown in red. The patient bed, which runs through the MRI scanner, restricts the available bore radius. The radius of the bore is of utmost importance since the monkey’s brain must fall exactly at the point known as isocenter. Isocenter is the point in the exact center of the magnet’s bore at which maximal signal is obtained. The loading of the chair into the magnet also needs to be taken into consideration since the entrance to the bore is roughly three feet off the ground. A consideration of weight must be kept in mind when selecting materials. The end weight of the chair must be considered to be the weight of the physical chair, plus the weight of any subsequent functional components as well as the weight of the monkey.
The cohort available for these studies is a group of rhesus macaques. Though only two monkeys will be chosen from the cohort for further training and use in visual experiments, the chair must be accommodative of variable monkey size. The cohort, currently residing in quarantine, ranges in size from 6.5-8.5kg in weight and 28-32” in length (in the prone position). The chair design and construction must be started while the monkeys are still in quarantine so that when they are released, the two monkeys chosen for training can make use of the primate chair during training.
The properties of materials were only considered for performance within the range of 10-35 degrees Celsius, as this is the complete temperature range to which the chair would ever potentially be exposed.All materials must be MR compatible and provide for optimal signal in imaging. Some materials provide better signal than others, but multiple factors must be taken into account during selection. It is very important that the materials are strong enough to contain the monkey and resist any force resulting from a non-behaving monkey. The material used for siding must provide isolation of the monkey from its environment. Waste from the monkey, which will likely collect at the base of the chair, must be prevented from leaking out into the environment of the bore. Finally, the chair must be translucent to allow for the monitoring of the monkey during the scan. A comparison of various known MR compatible materials was performed to evaluate the best material choice for the siding, boot piece and binding of the chair (Table 1). Though no budget restrictions were specified for this project, the cost of materials still needed to be considered, regardless of the material’s property qualifications.
Different materials, when placed near each other in an MR environment, have the potential to cause image artifacts due to their differing susceptibilities. Acquisitions of echo-planar MRI images reveal subtle differences in MR compatible materials. Chair siding necessitates the use of MR compatible materials, but head post apparatus and binding materials and screws closer to isocenter require a higher level of MR compatibility. The closer a material is to isocenter the more sensitive the image is to differing susceptibilities. A discussion of materials selection for head post apparatus and binding screws near isocenter is found in the Results and Discussion section of the paper.
Results & Discussion
Bore Constraints
Normally during an MRI scan the subject is loaded onto the patient bed and the bed is then electronically moved into the bore. As the bed begins to move into the scanner it briefly stops and marks what is known as the “landmark”. The bed continues to scroll into the bore and stops when the landmark reaches the center of the bore. In other words, the landmark is moved to the position y=0 along what would be the x-axis on planar coordinates relating isocenter. Before beginning chair design, the parameters of the bore were measured and recorded (Figure 7).The height between the base of the bore and isocenter, when measured with the patient bed inside the bore, proved to be suboptimal. It was determined that a monkey would not be comfortable, even in a sphinx position, with an available height of only 15cm.
By removing the patient bed, it was found that an extra 8cm of height was made available. Though this provides for the comfort of the monkey, the issue of setting a landmark arose. Without the patient bed in use, a landmark can not be set, and the isocenter not determined. A second patient bed was modified to include an acrylic plate near the end, flush with the bed’s surface, which had several holes drilled in it. Normally this patient bed is used for other experiments at the VUIIS, but its use in conjunction with the primate chair during rhesus scanning had the potential to be beneficial. The primate chair will be manually loaded into the bore. As a feature of the chair a lip will be constructed, that attaches to the end of the chair. The lip will be attached to the chair end with two bolts placed through vertical slots allowing the height of the lip to be adjusted. When set to the proper height, the lip will come off of the chair and extend onto the acrylic slab of the modified patient bed where it will be bolted to the bed using ¼”-20 brass screws. Once the chair is attached, a “fake” landmark point will be set a known distance from the monkey’s brain. Once the landmark is set, the scanner will automatically roll the patient bed into the bore a proper distance. The rolling of the patient bed provides sufficient force to push the chair into the proper position for the monkey’s head to be at isocenter, while avoiding cumbersome and awkward manual pushing. It also allows for a more accurate positioning of the monkey’s head at isocenter, as the scanner automatically pushes the bed (and chair) the required distance. The patient bed, now just outside of the bore, is adjacent to the monkey chair which resides inside the bore. The attachment of the chair to the bed also provides additional stabilization of the chair to avoid image artifacts resulting from primate movement.
Cohort Specifications
Accommodation of variable monkey size is accounted for in the design of multiple chair features. To further provide monkey comfort, an angled wedge will be placed under the chest to support the macaque sitting in a “sphinx” position, similar to the monkey positioning and wedge seen in Figure 4. The chest support will be physically independent from the chair to allow for unique placement under the chest of monkeys of varying size.
The neck plate, placed to enclose the font end of the tube, will be constructed of two pieces of polycarbonate. The neck plate will have a fixed front plate and a removable rear plate that screw together to allow for variable neck diameter. An example of a neck plate currently manufactured by Crist Instrument is seen in Figure 8. The neck plate helps to restrain the monkey without chafing. The rear plate will have two holes suitable for the monkey to put its hands through. The holes should allow for comfortable movement, while still restraining the monkey enough to disallow action such as touching of the head. By permitting the animal to use its hands, task response can be interactive beyond the tracking of eye movement. Further investigation is needed to determine the exact dimensions and parameters for integrating the neck apparatus.
The head apparatus must be very stable. The post that attaches the chair to the post in the monkey’s head must account for variable size of the monkey and position of the head in three dimensions. To accomplish this task a head apparatus was proposed with six degrees of freedom. Initial design ideas include a cylindrical rod running perpendicular to the monkey’s head post allowing for a site of attachment. The cylinder will be able to move toward and away from the monkey’s head as well as up and down through 3 rods attached to the chair. The cylinder will be attached to the rods in a manner that allows for angular rotation. No drawings are yet available to elucidate the concepts involved, though this should be the next step taken in the design process. After sketches are made and presented to the investigator, rigorous strength testing must be done to ensure that the six degrees of freedom don’t compromise the stability of the head apparatus. To further stabilize the head apparatus and address problems seen in the past, two vertical support rods connected to the horizontal head bar (cylinder) should be put in place. In the Pinsk model the support rods, though improved from Vanduffel’s design, did not produce optimal stability. To improve on the model the vertical supports should be set at an angle (to be determined), connected to the head bar and base of the chair at either end. This angular formation should provide more overall stability than the previous design using 90 degree angles. Figure 9, though it does not include the concept of head apparatus with six degrees of freedom, does show the relationship of the angled support rods to the head bar and base of the chair.