José L. Contreras-Vidal, Ph.D.
Assistant Professor
Behavioral and Computational Neuroscience
Research Foci
Lab Members
Collaborators
Research Opportunities: Undergraduate
Research Opportunities: Graduate
Research Opportunities: Postdoctoral
We use behavioral and computational neuroscience methods to study the neural mechanisms and computational principles underlying adaptive sensory-motor control in humans. Specifically, we study how the macro and microstructure of movement is altered by transient or permanent changes in the human brain due to environmental change, development, aging or disease, as compared to the young, mature, healthy population. Parkinson's disease is used as a window to study how the motor load is distributed in the basal ganglia -thalamo-cortical networks, and to investigate movement invariances that point out to the underlying mechanisms responsible for specific patterns of adaptation and movement. Similarly, studies involving patients with cerebellar dysfunction, such as olivopontocerebellar atrophy, inform us about the cerebellar contribution to motor learning and control. An important goal of our research is to understand how the brain reorganizes during across the life span and how it changes in response to novel environments. In this regard, we are studying how multiple internal models of the interaction between our body and the environment (e.g., during learning to use a new tool) are formed, and how these models are affected by disease (e.g., developmental coordination disorder, cerebellar dysfunction and Parkinson's disease).
Investigations aimed at determining the computational principles of adaptive sensory-motor control are two-fold: first, modeling the connectivity and neurophysiology of the brain allows us to integrate a large amount of biological data related to behavior and to provide a mechanistic account for movement invariances. Second, neural modeling allows us to examine in detail intervention procedures or predictions that can be tested experimentally. Importantly, top-down information provided by behavioral tasks is used to constrain the mathematical description of neural networks based on neuroscience data. Thus, the behavioral research is closely related to the computational neuroscience methods fostering the transfer of biological principles to robotics, bio-engineering and medicine.
 |
 |
||
| Figure 1. Schematic depicting the biologically-based architecture of a neural network model of reaching. The model includes basal ganglia, thalamic, cortical, spinal cord, and proprioceptor structures. |  |
||
 |
|||
 |
|||
|
|
|
|
We have identified
organizing principles across brain structures underlying the
learning, selection, planning, initiation and execution of willed
actions. In this regard, a neural network model of opponent
cerebellar learning for arm movement control (Contreras-Vidal,
Grossberg, & Bullock (1997) Learning
and Memory, 3:475-502) has shown that cerebellar learning
modifies velocity commands to produce phasic antagonist bursts
at interpositus nuclear cells (via learned disinhibition), whose
feedforward action overcomes inherent limitations of spinal
feedback control of tracking. Moreover, excitation of alpha
motoneuron pools, combined with inhibition of their Renshaw
cells by the cerebellum, facilitate movement initiation and
optimal execution. A similar biological principle has been discovered
in the basal ganglia-thalamo -cortical circuits (Contreras-Vidal
& Stelmach (1995) Biological
Cybernetics, 73:467-476), whereby the basal ganglia act
by opening normally closed gates via selective phasic removal
of tonic inhibition in the thalamus (the adaptive gating hypothesis).
Furthermore, this selective opening and switching of thalamo-cortical
pathways appears to occur in relation to expected or current
task requirements allowing real-time reconfiguration of movement
control pathways. Recent data support the view the basal ganglia
are also involved in cognition and other higher brain functions.
In this regard, we have addressed the computational roles of
dopamine cells in the learning of approach behavior (Contreras-Vidal
& Schultz (1999)
J Computational Neuroscience, 6:191-214). Specifically,
we proposed an adaptive neural network model of how basal ganglia
and prefrontal cortex interactions guide short and long-term
processing related to novelty, generalization, and discrimination
of appetitive and aversive stimuli during reward-related learning.
The model predicts that Parkinson's disease patients are impaired
in tasks involving trial-by-error learning.
Group Leader:
Dr. Jose Luis Contreras-Vidal (a.k.a. 'Pepe'), Engineer (Monterrey
Institute of Technology, 1987), M.S.E.E. (University of Colorado,
Boulder, 1990), Ph.D. in Cognitive and Neural Systems (Boston
University, 1994)
Faculty Research Associates:
Scott Kerick, Ph.D. in Kinesiology (University of Maryland, 2002):
- Current Research: Neurocognitive mechanisms underlying motor skill acquisition and performance. Age-related changes in cortical organization and kinematic variables associated with motor learning using independent component analysis and spectro-temporal analysis of high-density EEG data.
- Funding: NIH, Army Research Laboratory (ARL) - Human Research and Engineering Directorate (HRED)
- Email: kerick@wam.umd.edu
Florian Kagerer, Ph.D:
- Current Research: Development of visuomotor transformation in healthy children and in those with developmental coordination disorder (in collaboration with Prof. Jane Clark)
- Funding: NIH
- Email: florian.kagerer@utas.edu.au
Graduate Research Assistants:
Feng Rong, B.S. in Biology (Peking University, 07/1998):
- Current Program of Study: Ph.D. In Neuroscience and Cognitive Science Program, University of Maryland
- Funding: 1/2 NACS Fellowship; 1/2 GRA
- Current Research: I am working on a forward spherical volume conductor model on EEG. What I am intested in include both the forward and the inverse solution with this model. I am going to improve this model with less arbitrary source setting and integrate this model with other brain imaging techniques and a large-scale neural network for visuomotor learning in the near future
- Email address: rongfeng@glue.umd.edu
Shihua Wen, M.S., Dept. of Decision Science and Engineering System (Rensselear Polytechnic Institute, 08/2000), B.S., Dept. of Physics (Nanjing University, 06/2002):
- Current Program of Study: Ph.D. in Statistics in Dept. of Mathematics
- Funding: NIH (NIDCD)
- Current Research: a) Synthetic PET using a large-scale neural network model of visuomotor learning that includes the cerebellum, basal ganglia, premoter cortex, etc., which are brain areas related with arm or finger movement. Should we achieve this, it will greatly break through the physical scan time limit of current experimental PET study. b) Research on neural networks modeling to demonstrate the decison making process occurred in human brain when doing sample-match tasks, and show corresponding finger movement based on the designed delayed-match-to-sample decision signal. c) Research on adaptive neural networks movement control model to mimic Parkinson's disease patients, and investigate which area is the most significant area and how it causes Parkinson's disease, then give better treatment suggestions
- Email address: wen@math.umd.edu
Jin Bo, M.A. in Kinesiology (University of Maryland, 12/2002), B. Med. in Medicine (Shanghai Second Medical University, 07/1993):
- Current Program of Study: Ph.D. in Neuroscience and Cognitive Science Program at the University of Maryland (Co-mentored by Dr. Jane Clark)
- Funding: NACS fellowship and NIH
- Current research: 1). Development of visuo-motor coordination and adaptation in normally developing children; 2). Effect of Visuo-motor transformation on children's arm movement; 3) visuo-motor coordination in Developmental Coordination disorder children
- Email address: jbo@wam.umd.edu
Bruce Sweet, B.A. in Psychology and Philosophy/Religion
(
Western Maryland College, 1981), M.Div. (Brite Divinity School
(Texas Christian University, 1986), MSW in Clinical Social
Work (University of Maryland at Baltimore, 1993):
- Current Program: Ph.D. in NACS, University of Maryland
- Funding: NACS Summer Fellowship; KNES 1/2 RA, Lectureship, EDHD
- Current research: I am studying the Anterior Cingulate Cortex, and its role in error detection, volitional movement, and motor planning, using both EEG and MEG processes. Current endeavors include formal task analysis and the formulation of the research protocol for MEG study of error detection by the ACC
- Email address: b.swett@comcast.net
Danny Gold, undergraduate research assistant:
- Current Program: B.S. in Kinesiology, University of Maryland, KNES Honor's Program
- Funding: Howard Hughes Medical Fellowship, Summer Research Fellowship
- Current Research: I am investigating the sensory deficits of Parkinson's disease patients through the tendon-vibration paradigm. This consists of artificially stimulating muscle spindles (distorting proprioception) with vibrators placed on the biceps tendon and anterior deltoid of the dominant arm, and seeing how patients' disordered kinesthesia affects their performance in a center-out drawing task
- Email: golddr@wam.umd.edu
Sima Mistry, undergraduate research assistant:
- Current Program, B.S. Biology, University of Maryland Undergraduate Research Assistant Program
- Current Research: I am investigating how the brain learns multiple internal models of the environment. For this, I am testing young volunteers using a visuomotor adaptation task in which the relationship between hand movement and screen cursor feedback of movement is distorted
- Email: Tiredlion@aol.com
Past Students:
Sereniti Young, B.S. Kinesiology (University of Maryland, 2001):
- Supported by Howard Hughes Medical Fellowship for undergraduate research on aging
- Now at Baylor University Louise Herrington School of Nursing
- Email: sereniti13@excite.com
Ethan R. Buch, B.S., M.A in Kinesiology (University of Maryland, 2002):
- Supported by RA and the NIH
- Now at the Laboratory of Systems Neuroscience, NIMH
- Email: erbuch@codon.nih.gov
Moritz Grosse-Wentrup, B.S. in Automatic and Control Engineering at the University of Technology Munich:
- Funding: Institute of International Education (IIE)
- Email: mogrwe@eikon.tum.de
Ariel Prager, Undergraduate research assistant:
- Now at: Clinical Brain Disorders Branch, NIMH, 10 Center Drive, Room4s227 Bethesda
- Email: arielprager@hotmail.com
Jane Clark (University of Maryland), Ph.D.
Barry Horwitz (NIDCD), Ph.D.
Craig Carignan (Georgetown University), Sc.D.
David Poeppel (Univ of Maryland's MEG Center), Ph.D.
Prof. Juan Lopez-Coronado (Politechnic University of Murcia,Spain)
Neural networks for prehension, Robotics, Computer Vision
Dr.
Daniel Bullock (Boston Universit)y
Neural network dynamics, hardwriting, prehension, spinal cord,
movement control
Dr. Bradley D. Hatfield
(University of Maryland)
Psychophysiology, Human skilled performance, electroencephalography
(EEG), mental health, aspects of exercise
Research Opportunities: Undergraduate
Students receive credit or a note in their transcript through
the Undergraduate
Research Assistant Program (URAP), get paid interships
through the Summer
Research Program, or the Howard
Hughes Medical Fellowship Program
Typical Research Projects for Undergraduate Students
(summer, semester, or academic year):
How the brain learns visuo-motor transformations
for movement:
In visually guided arm movements, a perceived target direction
needs to be mapped into motor commands that drive the arm
in the intended spatial direction. This so-called visuo-motor
map needs to be updated continuously based on growth, environmental
changes, or aging. This line of research will explore the
effects of development, aging, brain disease or changes in
the environment on visuo-motor learning. Typical paradigms
would include changes in the relationship between hand movements
and the visual feedback of cursor movement on a computer screen.
Both normal and neurological populations (e.g., Parkinson's
disease) are studied.
Simulation of neural networks for movement control:
Several models for spinal cord, subcortical structures and
cortical brains areas involved in adaptive sensory-motor control
have been developed. Mathematical models that simulate the
behavioral operating characteristics of human and animal movement
have also been put forward. These models need to be tested,
refined or disposed. Engineering methods and/or control theory
can be used to evaluate these models.
Instrumentation and measurement:
Physiological, bio-mechanical or bio-potential variables need
to be measured using specialized sensors and circuits. Students
can participate in the development/refinement of new sensor
technology/circuits for measurement of variables in sensory-motor
research (e.g., grasping force, brain activity, or muscular
activity).
Research Opportunities: Graduate
Several lines of research within Behavioral and Computational
Motor Neurosciences, Complex Systems, Non-stationary Pattern
Recognition and Analysis, or Bio-robotics are available for
research work towards a Master’s or Ph.D. degree.
Cognitive-Motor Behavior in Parkinson’s Disease:
We use Parkinson’s disease as a window to study the
functions of motor, cognitive, and limbic networks of the
basal ganglia-thalamo-cortical circuits. Opportunities exist
for research in the learning, selection, planning, initiation,
and execution of actions using behavioral and/or computational
paradigms. Tools include EEG, MEG, Independent Component Analysis.
Disordered Motor Activity in Dystonia:
This line of research aims to understand the source of dystonia,
a disease characterized by disordered muscular, subcortical,
and cortical activity that results in muscular co-contraction
and overflow of EMG activity, slowness, prolonged muscle contractions,
abnormal posture of the affected body part(s), movement variability
and uncontrolled/involuntary movements. Opportunities for
research at the spinal cord, basal ganglia, or cortical levels
are available. Behavioral and/or computational paradigms can
be utilized to gain insight about the mechanisms underlying
dystonia.
Bio-robotics:
Students with a strong background in engineering and/or neuroscience
and who wish to undergo advanced training in the applied neuroscience/computational
neuroscience areas have opportunites for research/development
in the application of neural network theories/models to robotics.
Our project ``Systems Neuroscience and Engineering Research
for Antropomorphic Grasping and Handling ‘’ aims
to develop the next generation of artificial limbs/hands for
use in prosthetic devices, assembling lines, and robotic research
in general. In contrast to traditional robotic systems, our
work mimics both the bio-mechanical atttributes of the arm/hand
system and the neural network control algorithms that we believe
the brain uses to learn and control prehensile tasks.
Non-stationary Pattern Recognition and Analysis:
Biological data, including neurophysiological recordings,
sequence data, and behavioral data, may be characterized as
non-stationary spatio-temporal patterns of activity that need
to be analyzed, clustered, matched, filled-in, and correlated
with other brain or behavioral variables. Recent neural networks
for self-organizing pattern recognition provide useful tools
to aid in this task.
Research Opportunities: Postgraduate
Several opportunities for postdoctoral research exist within the Cognitive-Motor Behavior Laboratory in the Behavioral and Computational Motor Neuroscience Group.
Deadline: April 5, August 5, December 5, annually.
2 year support. Up to $40,000 per year.
Deadline: Sept. 1, annually.
3 year support. Up to $50,000 per year.
Deadline: Feb. 26, annually.
Deadline: December 30, annually.
