Functional MRI Core Facility

National Institute Of Mental Health

Investigators

Linked publications, trials & patents

Abstract

Resources: The Functional Magnetic Resonance Imaging Core Facility (FMRIF) manages five scanners: two GE 3T, one Siemens 3T Skyra, and two Siemens 7T scanners, designated 3TA, 3TB, 3TD, 7TA, and 7TB respectively. While FMRIF does not manage the NIAAA-managed Siemens 3T Prisma, NIMH and NINDS has two full days access per week, scheduled through FMRIF. Space Utilization: The FMRIF occupies approximately 5000 sq ft of space in the NIH Clinical Center, divided between the B1 level scanner bays, control rooms and electronics/machine rooms for 3TA/3TB, 3TD, and the two Siemens 7T scanners, (about 1800 sq ft, 1100 sq ft and 1300 sq ft respectively) and office space within the NMR center. On the first floor are the FMRIF/SFIM suite (approximately 800 sq ft total) for office space and shared conference space for all staff employed full-time by the facility. Staff: The current staff of the FMRIF includes the director, four staff scientists, seven radiological technologists, a Nurse Practitioner, and an administrator. One staff scientist is director of the Center for Multimodal Neuroimaging, managed within the FMRIF. Breakdown of senior staff general duties: Vinai Roopchansingh: Budget management, procurement, real time fMRI, neurofeedback platform, maintenance of pulse sequences, maintenance of backup systems, management of IT specialist. Andy Derbyshire: Management of technologist team and nurse practitioner. Gradient coil project, MR physics and pulse sequence consulting – GE specialty. Linqing Li: Transcranial Magnetic Stimulation (TMS) in scanner assistance, select software and equipment upgrading. Tyler Morgan: High field advanced pulse sequences, MRI reconstruction and processing, instrumentation. Pete Molfese: Directing the Center for Multimodal Neuroimaging, EEG/fMRI instrumentation, implementation, processing, and consulting. Investigators using FMRIF: In FY24, the FMRIF supported the research of about 30 Principal Investigators translating to about 300 researchers overall. About 60 research protocols make use of FMRIF scanners. Each scanner has scheduled operating hours of 105 hours per week. Relevant FMRIF Metrics: Papers published through use of the FMRIF: A strong measure of the utility of a core facility is the quantity and quality of scientific papers published by investigators using the facility. We have kept careful records of papers published and their corresponding citations. Since its inception in 2000 until July of 2025, 1,451 peer-reviewed publications from intramural investigators have used data acquired in the FMRIF core facility. The total number of papers is distributed among 837 papers from NIMH, 418 papers from NINDS, and 196 from the other institutes. These papers have been cited 132,397 times. During the current reporting period, from January 2024 until July 2025, a total of 83 peer-reviewed publications from intramural investigators have used data acquired in the FMRIF core facility through 89 IRB- approved protocols. The total number of papers is distributed among 42 papers from NIMH, 33 papers from NINDS, and 8 from the other institutes (NCCIH, NIAAA, NIBIB, and NHGRI). These papers have been cited a total of 162 times. Website: FMRIF: *In this annual report (12-month period) the FMRIF website has received 18,000 visitors. *We have the summer course up to date including last summer 2024: https://fmrif.nimh.nih.gov/index.php/SummerCourse *The summer course videos have been watched by more than 6,900 times in the last year period, for a total 599.7 hours of viewing. *XNAT, our data archive is currently storing all fMRI data since the core’s inception for 115TB of total data stored, including roughly 181,141 scans CMN: *CMN had 6,300 visitors Specific staff projects and support work: Vinai Roopchansingh: XNAT migration: Dr. Roopchansingh oversaw the migration of FMRIF’s XNAT instance (the primary DICOM warehouse resource for all human scanners in the NMR Center, where NIMH and NINDS users scan, including NMRF’s Terra 7T, NIAAA Prisma 3T, and NINDS’s HyperFine 0.055T human scanners) from an older hardware stack (a physical server hosting a 160 terabyte (TB) storage pool, which was already 75% filled) to new resources (a virtual machine, attached to a 250 TB storage pool). The new setup has a simpler and more robust configuration, making it easier to maintain and update in the future, as needs and opportunities arise. IT Infrastructure: Over the last year, Dr. Roopchansingh also worked with FMRIF’s systems’ analyst to complete its IT infrastructure reconfiguration to have high availability compute and storage resources, using almost exclusively open source and freely available software tools. The only significant costs that remain in FMRIF’s revamped IT infrastructure are hardware support contracts, and a license for backup software. In-scanner TMS and neurofeedback: After a long period of testing and development, in 2025, Dr Roopchansingh (along with another FMRIF scientist, Linqing Li) helped staff (Dr Lysianne Beynel and Neil Baker) from Dr Lisanby’s unit collect data from subjects in a functional MRI (FMRI) experiment leveraging both (FMRI) neurofeedback, with in-scanner simultaneous TMS (transcranial magnetic stimulation). This represents the culmination of an extended period of work to get IRB approval, the hardware approved by the NMR Center Safety committee, and working with the Section on Instrumentation to develop holders for the TMS hardware in the scanner bore. Data streaming optimization: As a result of this work, Dr Roopchansingh encountered data streaming performance issues in the MRI scanner platforms currently installed and operated by FMRIF. These issues placed constraints on the data collected for the above-mentioned experiment, and has caused Dr Roopchansingh to revisit his previous work and experience with ISMRMRD (an open source data format developed at NIH, to allow researchers more flexibility and options when reconstructing and processing these data). Dr Roopchansingh is now leveraging ISMRMRD, along with a prototype data streaming tool from Siemens Healthineers (FIRE), to build a converter to allow reconstructed image data from a FMRI experiment to be streamed directly to AFNI (software developed by the SSCC group at NIMH) in real-time, without having to go through the intermediate step of being converted to DICOM formatted data. This step seems to impose a non-significant overhead for real-time FMRI experiments, so Dr Roopchansingh will be pursuing this current path, as well as any others that may arise, to extract data from FMRIF’s MRI scanners in a more efficient and timely manner. Linqing Li: System Safety Testing and Evaluation Conducted a series of temperature measurements for EEG electrodes at 7T to ensure safety compliance. A full report has been drafted and shared with Pete for review; pending his feedback, the data will be submitted to the safety committee. In collaboration with Andy and Ruifeng (from Dr. Carlo Pierpaoli’s group - NIBIB), performed temperature and acoustic testing on the GE 3T head gradient coil using diffusion sequences. Initial results have been collected, and a formal report is in preparation. Identified and resolved imaging artifacts that appeared after gradient removal on the 3TB system. Updated reconnection procedures and recommended the use of a torque wrench to ensure secure and consistent gradient cable connections. Collaborative Technical Development Worked closely with Dr. Pierpaoli’s group on thermal and functional evaluation of the head gradient coil. Supported the deployment of the field camera system and acquired the first dataset from the GE-3TB scanner. Data analysis and validation are ongoing. Coordinated with Ruifeng on cross-platform field camera software setup and troubleshooting; addressed compatibility issues by initiating the acquisition of a dedicated PC laptop. Successfully ran and demonstrated the field camera software using data from Siemens, and GE systems. Scanner Maintenance and Operations Facilitated system-level maintenance for the 3TD, 7TA, and 7TB scanners, working with Siemens to address gradient coil failures and third-order shim amplifier malfunctions. Replacement parts were coordinated and scheduled for installation. Conducted sequence-specific temperature measurements at 7T to support ongoing system safety validation. Research Support and Experimentation Continued supporting the TMS-fMRI study with Lysianne, conducting four additional runs to optimize protocol design and maximize amygdala activation based on TSNR metrics. Planned and scheduled follow-up safety evaluations, including additional electrode temperature and acoustic testing. Scientific Contributions Designed, prepared, and presented an e-poster at the ISMRM conference, highlighting recent methodological developments, entitled “Bridging Neurovascular Coupling Models” This study proposed a unified framework integrating the classical Yablonskiy–Haacke and Davis models with concepts of scale-free dynamics, providing an empirical foundation for quantifying CMRO₂ under Self-organized criticality. This work is moving towards becoming ready for publication. Laurentius (Renzo) Huber: Mid-year, Dr. Huber has transitioned from the FMRIF to the Section on Functional Imaging Methods (SFIM), hence his work has been divided between FMRIF development and SFIM research. FRISGO: FuzzyRipple-free Imaging with Short-term Gradient Optimization High resolution functional imaging can be limited by artifacts of short term gradient field imperfections. Dr. Huber has developed a echo planar acquisition approach FRISGO (Fuzzy Ripple-free Imaging with Short-term Gradient Optimization). This method allows fast fMRI with laminar resolution and whole brain coverage. This method was presented at international research conferences and has increased the number of users of Dr. Huber’s method by 30% from 47 to 62 within the last two months. Parallel transmit at 7T Dr. Huber has led a consortium of 28 global MRI research labs investigating the performance and utility of so-called pTx coils with sTx coils. These new coils are using clinically approved technology to homogenize the local distribution of MR signals across the brain. However, little was known about their practical performance for the purpose of functional imaging. This consortium could clarify the capabilities and challenges of upper and lower brain areas for the respective coils. Community service for high resolution fMRI As part of his official duty, Dr Huber has organized an international workshop of 200-300 in-person attendees about fMRI at UHF in Annapolis, MD. Dr. Huber serves as an advisor, advisory committee member, and steering committee member in 4 ongoing NIH grants between Oct 2024-June 2025. Dr. Huber left the intramural program in June 2025. Fast structural scanning with inversion-recovery EPI Aside from daily operations, each technologist specializes in their own applications, including implant safety, user training, patient recruitment, etc. Mr Chung’s specialty is MR-physics. He has led a project to implement testing and disseminating a novel more efficient acquisition approach of structural data acquisition that is usually accompanied with spatially distorted fictional scanning. This method – termed T1234 – uses a T1-weighted acquisition with two inversion times and a 3D-EPI readout while combining four combinations of spatial encoding polarities. This allows to reduce T1 mapping protocols A. Tyler Morgan: Dr. Morgan’s role in the Core Facility is to develop, validate, and apply acquisition and analysis tools for high spatial and temporal resolution fMRI. She has been facilitating and supporting neuroscience-focused applications of high-resolution fMRI across institutes. Ongoing Projects: The development and testing of non-selective MRI sequences to investigate the possibility of directly recording neural signals from fMRI data. This work aims to understand if millisecond-resolution neural signals can be detected. She has been working on development of functional sodium imaging in human brain, as recent reports at scientific meetings indicate that the T2* of sodium changes rapidly with sodium efflux into neurons during their action potential. Development of deep sampling protocols to measure layer-specific brain activity using high-resolution fMRI. This work aims to characterize layer signals at high signal-to-noise to better understand their inter-laminar and inter-area connectivity. In collaboration with Harri Piitulainen (University of Jyväskylä) and Renzo Huber (MGH), measuring layer-specific responses to active and passive finger movements to probe the neural mechanisms of proprioception. In collaboration with Max Reisenhuber (Georgetown), examining one- or few-shot learning in human Frontal and Temporal Cortices. Functionally time-resolved fMRI is a novel method that allows researchers to record the timing of experimental and physiological processes during MRI acquisition. These data are retrospectively sorted and binned at the k-space to create images of brain activity and physiologic processes. This approach differs from traditional fMRI which works with reconstructed images, limiting temporal resolution to that of individual image acquisition. John (Andy) Derbyshire: Dr. Derbyshire is working with Ruifeng Dong (a postdoc in Dr. Pierpaoli’s NIBIB Laboratory) on the insertable head-gradient system. A safety report to the NMR Center Safety Committee on using the device with human subjects is in preparation addressing issues with dB/dt (slew rate), SAR (internal RF coil), mechanical, electrical and acoustic safety. Dr. Derbyshire is developing balanced SSFP-based pulse sequence to exploit the advantages provided by the head gradient to provide high SNR imaging while limiting the SAR exposure to safe levels. The sequence is inspired by the MCDESPOT sequence for Dr. Berman’s group (see below) and, in part, seeks to overcome some of the SAR issues encountered with that sequence. Dr. Derbyshire is assisting Dr. Ruifeng Dong (Pierpaoli Lab, NIBIB) to quantify diffusion measurements. The project measures both the diffusion and imaging gradient waveforms (e.g. using the Skope Field Camera) during the MR imaging process. Inclusion of the imaging gradient contributions, eddy current effects, and subtle differences between specific aspects of different vendors implementations of the dMRI sequences explains the majority of discrepancies seen between quantitative diffusion measurements made on nominally equivalent systems from the different vendors (GE, Siemens, Philips). Accounting for eddy currents during the image formation process helps to significantly improve the image quality (reduce geometric distortions and ghosting) in quantitative diffusion MR. Dr. Derbyshire continues to support the provision of MR pulse sequence and reconstruction development environments for users of the FMRIF and other groups who wish to develop their own research pulse sequences. Dr. Derbyshire is the day-to-day coordinator for the FMRIF team MRI technologists and acts as an interface between the research groups and the technologists to facilitate the introduction new scanning protocols. Dr. Derbyshire is a member of the NMR Center Safety Committee which reviews submissions for research coils and other equipment to be used in combination with the MRI systems. Dr. Derbyshire is also a member of the NMR Center Implant Working group which identifies and evaluates the safety of scanning new implants, particularly at high field (7T) where there may not be any explicit manufacturer guidance. Dr. Derbyshire serves as member of the NMR Center Safety Class team including reviewing, developing and presenting the MR Safety Class required by all users of the NIH NMR Center and employees and contractors working in the NMR Center. The Safety Class team is finalizing an on-line refresher course to facilitate user’s re-certification. Dr. Derbyshire continues to provide support for Dr. Berman’s team by maintaining the MCDESPOT suite of pulse sequences. Dr. Derbyshire provides support to Dr. Daniel Reich (NINDS) for the Multiple Sclerosis Cooperative (NAIMS) by maintaining and distributing 3D EPI pulse sequence to identify the central vein in MS lesions. Peter Molfese: The Center for Multimodal Neuroimaging (CMN) seeks to bridge the gaps between different neuroimaging modalities. In recent years, it has accomplished this through scientific consulting on new and continued projects, as well as maintaining three EEG systems. Of the EEG systems and peripherals CMN supports, two EEG systems are portable and able to interconnect with three NIH-owned 3T MRI scanners. The CMN continues to grow a user-base amongst scientific groups within the IRP. During this fiscal year, two groups within NIMH have begun collecting data using EEG equipment maintained by the CMN after completing training in use and safety of simultaneous EEG equipment. This adds to the existing three groups that continue to use CMN resources in data collection, analysis, and manuscript writing. With the increase in EEG users, the CMN now allocates more time towards the maintenance of EEG systems (e.g. net repair, computer upgrades) and upgrading peripherals (e.g. stimulus computers, eye trackers, analysis workstations). The CMN continues to work towards safety certification of the high-density EEG system at 7T MRI. Scientific Projects: Sleep-Cognition – Project collaboration with Walter Reed. Seven subjects in sleep restriction protocol completed between 3 and 5 visits encompassing baseline, mid-sleep restriction, end sleep restriction, acute recovery, and long-term recovery. Analysis in progress. Naturalistic MEG – Project collaboration with Elizabeth Ballard, Jessca Gilbert, and Carlos Zarate. Analysis nearly finished and manuscript is being written. Collaboration with University of Delaware on overnight sleep consolidation memory with EEG rest measures and fMRI outcomes. Follow up to recently published paper (Earle, Molfese, & Myers, 2025).