Peter Bandettini, Ph.D.
Section on Functional Imaging Methods, Laboratory of Brain and Cognition (LBC) / Functional Magnetic Resonance Imaging Core Facility (FMRIF)
Magnuson Clinical Center (Building 10), Room 1D80B
10 Center Drive
Bethesda, MD 20814
The goal of the Section on Functional Imaging Methods (SFIM) is to better understand and more efficiently extract neuronally and, ultimately, clinically-relevant signal from MR images and fMRI time series. Most of our work since 2007 reflects this goal, as signal interpretation research and processing method development have constituted the bulk of the 47 papers and 89 abstracts produced by SFIM during this time period.
Specifically, since 2007, we have been working towards obtaining a deeper understanding of the resting state and activation-based signal change characteristics, and designing better methods to extract neuronal information from noise and irrelevant signal. This brief description highlights our efforts on: 1. Spontaneous fluctuations; 2. Activation dynamics, patterns, and mechanisms; and 3. Anatomic MRI, perfusion, and fMRI calibration.
1. Spontaneous fluctuations: Since 2007, the field of “resting state fMRI,” (rs-fMRI), also known as the study of “spontaneous fluctuations,” and a cornerstone of the field of “functional connectivity,” has exploded. Just under 800 papers are projected to be published on rs-fMRI in 2012 - about a third of all projected fMRI papers for this year.
This explosive rs-fMRI growth has resulted from a convergence of the availability of high MRI scanner stability and sophisticated processing methods, as well as a growing number of repeatable and relevant findings. Nevertheless, fundamental questions related to the scanning and processing methodology, neuronal and behavioral correlates, as well as the variability of rs-fMRI remain. In addition, complete and objective elimination of nuisance fluctuations remains an elusive goal. Our group is uniquely positioned to address these methodological and interpretive issues. We have focused much of our effort towards improving not only how resting state data can be acquired and processed but how it can be interpreted. We are also exploring what further information can be extracted and used. We plan to continue with this rich line of research – pursuing a deeper understanding of how correlated networks change across time scales as well as how networks can be classified and related to populations and to specific characteristics of individuals.
We have data demonstrating a combined multi-echo pulse sequence and ICA-based processing stream for robust and automatic identification of relevant ICA components and for reduction of non-blood oxygenation level dependent (BOLD) fluctuations. From this foundation, we have developed a rs-fMRI based clustering of the entire brain into functional segments, as well as a method to minimize ubiquitous motion-related confounds in resting state data. We have also characterized, using a sliding window, the temporal variability of rs-fMRI correlation data. Lastly, we are developing methods for classifying time-varying mental states as a function of network correlation structure.
2: Activation dynamics, patterns, and mechanisms: A cornerstone of SFIM research is the improved understanding and use of characteristics of the task-activated hemodynamic response. For over thirteen years, we have been probing the dynamics, spatial patterns, behavioral correlation, and variability of this response. Perhaps our most exciting and novel finding in this theme has been the discovery that even with a simple task, the entire brain shows evoked regionally-specific and uniquely shaped hemodynamic responses(Gonzalez-Castillo, Saad et al. 2012). To obtain these results we averaged responses over nine hours from individual subjects performing a simple discrimination tasks, then processed the data with no assumptions about hemodynamic shape. Clustering of the responses showed clear responses throughout the brain. Also within this theme, we demonstrate that reduction of the MRI flip angle - having several advantages over a higher flip angle - does not result in a reduction in temporal signal to noise if physiologic noise dominates(Gonzalez-Castillo, Roopchansingh et al. 2011).
We also report here on our ongoing efforts towards better understanding the underlying mechanisms behind and limits of multivariate analysis – considering each voxel as part of a pattern of activity across relevant regions of the brain. Using a multivariate classification approach, we have been able to infer sub-TR (100 ms) and sub-voxel (ocular dominance column) relative activation sequences, and have been able to decode “yes” and “no” answers in subjects - both of these at about 90% accuracy.
3. Anatomic MRI, Perfusion, and fMRI Calibration: Our group also values the importance of developing methods that are complementary to rs-fMRI and fMRI. A new area of research has emerged that is based on the measurement of anatomic MRI changes in individuals over short periods of time. We have established a processing pipeline for assessing changes in anatomy with exercise and learning (Thomas, Marrett et al. 2009; Thomas, Dennis et al. 2012). We continue to develop perfusion imaging methods, working closely with fMRI core facility. We have also developed a high resolution, off-resonance corrected arterial spin labeling (ASL) approach at 7T resulting in perhaps the highest fidelity ASL-based perfusion images yet produced (Luh, Talagala et al. 2012). Regarding hemodynamic calibration, we show the utility of using the Valsalva maneuver. This finding may lead to implementation of a more simple and repeatable global calibration stress than that of hypercapnia.
Since 2007 our group has also produced a body of work that has involved a wide range of applications of cutting edge methods, but due to space limitations, this research will not be described. These topics include anatomic differences with autism (Misaki, Wallace et al. 2012), non-neuronal related release from adaptation (Mur, Ruff et al. 2010), single stimuli activation profiles (Mur, Ruff et al. 2012), movement-related theta rhythm assessment (Kaplan, Doeller et al. 2012), unconditioned responses and implicit memory (Knight, Waters et al. 2009; Knight, Waters et al. 2010), effects of task-correlated breathing (Birn, Murphy et al. 2009), group differences in connectivity (Jones, Bandettini et al. 2010), verbal fluency (Birn, Kenworthy et al. 2010), and regions differentially associated with task difficulty and decision making (Ruff, Marrett et al. 2010). In the future, we plan to move more more aggressively towards development of individual assessment methods. We believe that for fMRI to become more relevant to the mission of NIMH, it has to aid in assessing psychiatric and neurologic disorders on an individual basis – in much the same way that anatomic MRI is effective in identifying and monitoring an individual’s anatomic brain pathology. We are well-positioned to take this goal on, as we are working at 7T with novel pulse sequences and processing strategies.
Dr. Bandettini is Chief of the Section on Functional Imaging Methods in the Laboratory of Brain and Cognition of the Intramural Research Program, National Institute of Mental Health, National Institutes of Health. He is also the director of the Functional MRI Core Facility which provides functional MRI support to NIMH, NINDS and several other institutes at the NIMH. He received his B.S. in Physics from Marquette University in 1989 and his Ph.D. in Biophysics from the Medical College of Wisconsin in 1994, where he pioneered the development ofFunctional MRI (fMRI). During his post doc at the Massachusetts General Hospital, he continued his work on cutting edge fMRI methods development. In 1999, he joined NIMH as a tenure track investigator and a core facility director. In 2002, he received the Wiley Young Investigator Award. In 2001 and 2007 he received the NIMH Director’s Merit Award for his work in creating and maintaining the FMRIF. In 2012, he received the NIMH Outstanding Mentor Award. In 2006, he was president of the Organization for Human Brain Mapping. He is Editor in Chief of the journal, NeuroImage. He has published over 110 peer-reviewed papers, 18 book chapters, 1 book, has 1 patent and has presented over 250 invited lectures. His work has been cited over 10,000 times.
Huber L, Handwerker DA, Jangraw DC, Chen G, Hall A, Stüber C, Gonzalez-Castillo J, Ivanov D, Marrett S, Guidi M, Goense J, Poser BA, Bandettini PA. High-Resolution CBV-fMRI Allows Mapping of Laminar Activity and Connectivity of Cortical Input and Output in Human M1. Neuron. 2017;96(6):1253-1263.e7.
Gonzalez-Castillo J, Hoy CW, Handwerker DA, Robinson ME, Buchanan LC, Saad ZS, Bandettini PA. Tracking ongoing cognition in individuals using brief, whole-brain functional connectivity patterns. Proc Natl Acad Sci U S A. 2015;112(28):8762-7.
Finn ES, Corlett PR, Chen G, Bandettini PA, Constable RT. Trait paranoia shapes inter-subject synchrony in brain activity during an ambiguous social narrative. Nat Commun. 2018;9(1):2043.
Kundu P, Brenowitz ND, Voon V, Worbe Y, Vértes PE, Inati SJ, Saad ZS, Bandettini PA, Bullmore ET. Integrated strategy for improving functional connectivity mapping using multiecho fMRI. Proc Natl Acad Sci U S A. 2013;110(40):16187-92.
Huber L, Ivanov D, Handwerker DA, Marrett S, Guidi M, Uludağ K, Bandettini PA, Poser BA. Techniques for blood volume fMRI with VASO: From low-resolution mapping towards sub-millimeter layer-dependent applications. Neuroimage. 2018;164:131-143.
Related Scientific Focus Areas
Biomedical Engineering and Biophysics
Social and Behavioral Sciences
This page was last updated on August 15th, 2018