People - Postdocs

##       Natalie Adams

  My interests lie broadly in understanding how the brain represents perceptual information and why
  this may differ in pathological states related to autism and schizophrenia. I am tackling this using
  in vitro electrophysiological techniques in collaboration with mathematical modellers to look at
  mechanisms of network rhythm generation, modulation and interaction in the anterior cingulate
  region of prefrontal cortex.

          Currently, my research is focussed on exploring aspects of frontal cortical function that facilitate
  learning of sequences of sensory events (e.g. Siegel, M., et al 2009). I am interested in how precise
  spike timing in individual neurons or small sub-populations relates to the local field oscillation as a
  marker of the overall average of event timing relevant to a given stimulus.

          So far my work has revealed that many neurons in the anterior cingulate cortex possess the ability
  to intrinsically oscillate at sub-threshold levels. With varying degrees of tonic excitation these sub-
  threshold oscillations (STOs) exist at a variety of frequencies up to c.30Hz. The anterior cingulate
  cortex is known to have multiple mechanisms for the generation of gamma rhythms associated with
  cognitive function and I will look at how these rhythms interact with cellular STOs to affect spike/
  phase relationships and perhaps code for sequences: The working hypothesis is that assemblies
  of cells receiving higher levels of excitation increase the drive to the kinetics responsible for STOs,
  ultimately leading to a spike phase-advance on each period of field gamma.

          This work could uncover a substrate for the stable, computationally useful, temporal separation of
  concurrently active sensory representations.

##       Mikio Aoi

      My work has been focused on two aspects of the analysis of neural rhythms.  The first project, which I have been working on with Uri Eden, Mark Kramer, and Kyle Lepage, has focused on the spectral analysis of spike trains including coherence between signals when at least one of those signals is a spike train.  I use point process theory to derive properties of point process spectra and the estimators, and try to understand how those properties may help or hinder our understanding of the underlying neural system.  The second project is in collaboration with Timothy Gardner and Uri Eden in which we are investigating methods of multi-scale time-frequency analysis based on an object-based signal representation.  This method will allow us to extract signal information using multiple times scales simultaneously. This method may help to construct sharper spectral representations than are currently possible and we believe this operation may help to understand the phenomenology of human auditory perception.      

##       Justin Kinney

      Recording of neuronal spiking activity in distributed brain circuits
  requires a scalable design for massively parallel recording of
  extracellular field potentials. We are inventing such a system and
  implementing a proof-of-concept instantiation.  In this system,
  multi-electrode arrays are used, which minimize tissue damage and help
  with spike sorting, and time domain multiplexing of analog field
  potential acquisition reduces interconnect. Channel data is then
  relayed to a custom-designed terabyte capacity storage network via
  custom digital circuitry. The storage network is designed to enable
  neural data to be analyzed in flexible ways, including the evaluation
  of spike sorting methods.

          On the technical side, I am solely responsible for the design and implementation of the ethernet network and high-speed data storage software. In addition, I provide leadership to the project by staying well-versed in all aspects of the system design and maintaining open lines of communication between all technology developers, as well as organizing and documenting the design of the system.

##       Jung Lee

      Jung has been working with Kopell and Whittington on several modeling projects.  The main one concerns the effects of top-down beta rhythms on attention; Jung showed that such signals resonate with cells in the deep cortical layers, producing gain control and more gamma rhythms in the superficial layers; a paper is almost complete.  This work is highly relevant to work done by Miller on top-down attention, and further collaborations are planned.  The work also has relevance to aspects of schizophrenia, and conversations are beginning with the group of Kevin Spencer.  A second project concerns multiple inhibitory cell types in the rat auditory cortex.  

##       Kyle Lepage

      Kyle has been one of the most active members of the data analyis group. In the past year, he has been involved in CRC related activity involving three main subjects and two more tertiary ones. One primary project was a collaboration with the Kramer, Eden and Desimone groups on spike-field association (statistical procedures used to infer relations between a rhythm in a time series, such as a local field potential recording, and the firing activity of single neuron.  Mikio Aoi is also involved. There is now a preprint.   A second major project is a collaboration with the Eichenbaum and Eden groups on cells that measure time.  More technically, the project deals with the development of statistical procedures to separate the relative influence of covariates of interest such as time and rodent position upon neural activity.  There are two papers and several popular press articles about this work.  The third major project is a collaboration with the Kramer lab, also involving postdoc ShiNung Ching; it is motivated by techniques used in MEG and EEG experiments to find functionally connected networks, as in the Human Connectome. This work deals with principled estimation of the statistical connectivity between nodes in an evoked network. In this paradigm a stimulus is repeatedly applied to network nodes, one at a time, and evoked activity at nodes is used to infer a statistical relation between node activity. There is a preprint.  A smaller project with the group of Shinn-Cunningham concerns MEG eigensource. In this work local bias in MEG source estimates is traded for decreased non-spatially local bias due to unavoidable inverse-problem source localization limitation.  A final project, with Kramer, deals with removing bias in EEG measurements due to activity present on either an EEG reference electrode or present in a "re-referencing method"

##       Martin Luessi

##       Morteza Moazami

      Morteza is a postdoc in the lab of Miller.  He is interested in the functional circuitry for memory and context formation between and within the prefrontal cortex (PFC) and the medial temporal lobe (MTL).  PFC neurons reflect the associative relations between stimuli, task instructions, behavioral responses, rewards, etc. Interestingly, MTL neurons show similar properties. Neurophysiological studies have been focused on either the MTL or PFC and the interaction between the MTL and PFC is still unclear. He will simultaneously record - with many electrodes- from the PFC and MTL areas while monkey perform a task that temporally separates neuronal information related to context, sample, and recall of the correct choice. He will investigate the modulation of oscillatory neuronal dynamics between and within the PFC and MTL, during different stage of the task. The second project of Morteza concerns the oscillatory neuronal dynamics of categorization in the PFC.   Modulations of neuronal oscillations in the PFC with cognitive demands may regulate whether PFC neurons function as multitaskers. The data from these projects are essential to understanding central questions about the roles of rhythms in cognition, and will provide the basis for modeling efforts.  In addition, Morteza helps to run the physiology working group.

##       Lara Rangel

      Lara is a new postdoc in the lab of Eichebaum (BU). Her work was described above. She interacts frequently with members of the statistics and modeling groups (Eden, Kramer, Kopell).  She also talks frequently with members of other CRC labs (Boyden lab, MIT : Annabelle Singer; Wilson lab, MIT : Greg Hale, Sage Chen, and Stuart Layton) to compare data, methodology, and analysis techniques.  She is planning further interactions with Omar Ahmed of the Cash lab (MGH) on hippocampal oscillatory activity in humans, the Miller lab (MIT) on beta frequency oscillations, and the Jiamin Zhuo and Nick James of the Han lab (BU), on dentate gyrus function and beta rhythms, respectively.   She has also interacted with Whittington at CRC events.  She speaks frequently with postdocs Annabelle Singer, Justin Kinney, Omar Ahmed, and Kyle Lepage as well as graduate students Caroline Moore-Kochlacs, Greg Hale, and all the members of the Eichenbaum lab.  Lara has already written a grant proposal based on this work: NIH F-32 Individual Postdoctoral Research Grant

##       Wei Tang

      Wei studies large-scale networks in the resting human brain with data
  from non-invasive imaging techniques. Under the hypothesis that
  particular sets of brain regions interact with each other to maintain
  an active yet stable intrinsic state, the goal of her work is to
  uncover both the structure and dynamics of such intrinsic networks, in
  the hope that knowledge of the resting state will lead to further
  understanding of how neural electrophysiology gives rise to cognitive
  phenomena.

       Her current project involves collaborations between several
  laboratories. With MEG data acquired by Stufflebeam’s group, Wei and
  Steve are looking at seed-based Granger-causality maps, assessing
  their spatiotemporal and spectral properties, to explore their
  relationship with the proposed default-mode hypothesis. They will
  later extend the analysis to task data from the same subjects and see
  how the networks change undergoing different cognitive processes.
  Meanwhile, supervised by Matti Hamalainen and Uri Eden, Wei interacts
  with Patrick Purdon’s group at the Martinos Center, developing a
  state-space model based approach to identify the full
  source-connectivity matrix of the MEG signal and monitor its change
  over time. Efforts are being made to advance the methodology dealing
  with high-dimensionality of the data and make the full-network
  tractable. The third collaboration is with Mark Kramer, aiming at
  finding plausible biophysical models that can explain the observed
  network properties. This is an open area of exploration and may serve
  further modeling studies on brain disease such as epilepsy. Results
  from these collaborations together may provide a comprehensive picture
  of how the brain works at different levels.