Facilities and Resources

Facilities and Resources

The Stanford Center for Cognitive and Neurobiological Imaging (CNI, http://cni.stanford.edu/), a shared facility dedicated to research and teaching, opened in 2011. The center director is Prof. Brian Wandell, Dr. Adam Kerr as Research Director, and Dr. Laima Baltusis as the Facility Manager. The CNI supports scientific investigations into the brain that make rigorous connections between neuroscience and society. The CNI mission is to (1) Support neuroscience discovery for enhancing society, (2) Develop and disseminate cognitive and neurobiological imaging methods and (3) Create a structured, safe, and innovative environment for human neuroscience research.

The CNI is equipped with a 3T GE Ultra-High Performance MR system and and EGI 256-channel MR-compatible EEG system. It is maintained by a dedicated staff who oversee daily operations and maintenance. The CNI is a GE research partner and is supported by an MR Physicist from GE who is partially dedicated to the center. Further MR Physics support is provided by Dr. Hua Wu, a full-time MR Physicist on staff at the CNI, as well as collaborations with the Stanford Magnetic Resonance Systems Research Lab (MRSRL). In addition to the basic imaging equipment, the CNI has a wide array of state-of-the-art peripheral equipment.

The system is equipped with several standard coils from GE: a 48-channel head coil, a 28-channel head and neck coil, and several coils by NOVA Medical Inc including a 32 channel brain array (dimensions: 18.5 cm wide, 21 cm high) for collection of MRI images of the whole brain, and a 16 channel visual array for fMRI of visual cortex (dimensions: 18.5 wide, 10 cm high).

Visual display of stimuli during scan is achieved through one of two systems: a 47" 3D display by Resonance Technology which provides the ability to easily present different stimuli to each eye, and a data projector that provides a brighter stimulus with a larger field of view than the possible with the LCD displays. Eye movements can be tracked using an MR-compatible SR Research EyeLink 1000 with remote optics.

Auditory stimuli can be presented by low-profile MR-compatible headphones from Resonance Technology or through a custom system that uses bone-conduction transducers.

For collecting subject responses, we have a modular response box system (fORP 932) from Current Designs. This system allows users to swap out various response devices. The output from any of these devices is available from the fORP box via USB keyboard emulation. We have confirmed that the fORP 932 USB interface is polled at 1kHz, allowing response time measurement with 1 millisecond resolution.The response devices that we currently have include a scroll-wheel device, bimanual button boxes, a 5-button response pad, a joystick, and a 4-button stick-style response device.

MRI Simulator (mock scanner): The MRI simulator (Psychology Software Tools, Inc.) is designed to parallel the experience an individual will encounter during an actual MRI scan. The simulator consists of a 12-foot mock scanner with 6-foot tapered bore, head coil, head stabilizer system, visual and auditory presentation systems, and equipment for monitoring the subject's response to the training procedure. Two speakers are placed within the bore in order to deliver simulated MR gradient sounds to the subject. Behavior analysis and therapy techniques are used to counter conditioned fear or anxiety experienced in response to the imaging equipment and procedures. In brief, this process involves careful control of the salient stimuli in the imaging environment, systematic gradual exposure to the equipment, personnel and sensations involved in image acquisition. The simulator apparatus is housed in a separate room within the CNI facility.

The simulator allows researchers to train subjects to inhibit body motion during scanning. Typically, the subject will to listen to music, hear a story or watch a movie during the training. This "entertainment" serves as both distraction and immediate reinforcement for inhibiting movement. If the subject exceeds a pre-set movement criterion, the entertainment is automatically interrupted for a brief interval to provide a contingent consequence for body motion. The entertainment resumes when movement is again inhibited below the established criterion. The movement criterion can be adjusted according to the subject's current ability and gradually made more stringent as training progresses. The entire training process usually can be accomplished in one to three 30 minute sessions.

Data Collection: Data collected at the CNI are archived in a permanent database using software provided by Flywheel. Flywheel stores and organizes the raw data from the MR scanner as well as physiological measurements from the subject (synchronized with the MR data). Flywheel allows data to be shared within the CNI community, in accordance with regulations set forth by the Stanford IRB and the usage policies that apply to each investigator. Flywheel is securely hosted within a Google Cloud Project, provisioned for this specific use by Stanford's IT. Flywheel data are stored within a multi-regional object storage bucket to safeguard against natural disaster. The Flywheel database, as well as the data objects, are backed up hourly. User's authenticate to Flywheel using their Stanford University credentials. Authentication is handled via Stanford Login, backed by a Stanford/Google SAML 2.0 integration, which works across organizations and supports federation as well as multi-factor authentication protocols.

Data Processing: Data preprocessing is handled by Flywheel within its scalable cloud compute platform. Flywheel provides several pre-packaged processing algorithms (Gears), some of which were custom-developed at CNI. Gears are configured to run automatically on CNI data as they are ingested into Flywheel. These algorithms include extraction of metadata, data classification, DICOM-to-NIfTI conversion, reconstruction of raw k-space data, calculation of regressors from physiological recordings, and generation of automated quality assurance reports. The Flywheel platform is extensible and several other algorithms can be run on data once they are in the platform. Data can be downloaded from Flywheel using the Flywheel CLI, SDKs, or user-interface.

Skope system: The CNI is also equipped with a Skope Dynamic Field Camera that can be used by researchers to accurately measure the time-varying magnetic fields. The system includes 16 field probes and software that fits up to 3rd order spherical harmonics of the magnetic field with one-microsecond temporal resolution. The CNI also has a license to the Skope software package that uses these measurements for iterative image reconstruction and can support a wide number of acquisition strategies including SMS, MUSE, non-Cartesian and inplane accelerated scans. Use of the Skope requires researchers to be competent with making modifications to their MRI sequences of interest and requires special training. Please contact the CNI Research Director for more information.