Wednesday, June 13, 2012: 1:30 PM
- 3:30 PM
Room: Plenary Hall, Level 4
Wednesday, June 13 & Thursday, June 14
Randall Barbour1, Randall Andronica1, S. S. Barbour2, Harry Graber1, Daniel Lee1, Pflieger Mark3, J. D. Nicoles3, Yaling Pei2, Douglas Pfeil1, Christoph Schmitz4, Anandita Tyagi2, Yong Xu1
1SUNY Downstate Medical Center, Brooklyn, NY, 2NIRx Medical Technologies LLC, Glen Head, NY, 3Source Signal Imaging Inc, San Diego, CA, 44NIRx Medizintechnik GmbH, Berlin, Germany
Randall Barbour, Ph. D.
SUNY Downstate Medical Center
Near infrared spectroscopy (NIRS) and electrical
encephalography (EEG) are complementary technologies that support
exploration of the principal hemodynamic and bioelectric phenomenologies
associated with neuroactivation. To harness the full potential of these
methods, it is important that elements of data collection, information
extraction and feature validation be attended to in a comprehensive
manner. While direct measures are well developed and are easily
appreciated, extraction and validation of nonobservable features, such
as effective connectivity, is considerably more challenging. Ideally,
derived information would be mapped to a common brain space, while
validation would be supported by resources that mimic the full spectrum
of expected behaviors for both sensing domains, as least on a
macroscopic level. Here we report on efforts to extend capabilities
pertaining to extraction and validation of information derived from
measures of these domains. Building on previously reported work ,
validation is accomplished using a longitudinally stabilized, biomimetic
head form containing programmable source elements that can be precisely
mimic simple or complex hemodynamic and bioelectric behaviors.
Anthropomorphic Dynamic Phantom: The earlier approach 
was extended by introducing a "brain" (Fig. 1A) composed of a
hydrogel-based biopolymer with saline added to mimic impedances typical
of real tissue. Stabilizers were included to inhibit bacterial and mold
growth, and TiO2 and India Ink are added to provide physiologically
plausible optical coefficients. The embedded source array (Fig. 1B)
comprises electrochromic cells as optical modulators to mimic the
hemoglobin signal and dipoles for generation of bioelectric signals.
and Mapping Environment: For processing of NIRS data, we have
introduced NAVI [2,3], a MATLAB-based environment that supports data
transformations common to evaluation of bioelectric and hemodynamic
studies and atlas-based parametric mapping with full 3D tomographic
capabilities. The package includes modules for image formation, display
and analysis and a number of utilities modeled principally after
strategies supported by SPM8 [4,5]: GLM-based parametric mapping for
detected hemodynamic response functions; atlas-based mapping of image
findings onto identified brain regions, anatomical labeling
functionality; and examination of effective connectivity via strategies
such as dynamic causal modeling (DCM) .
The analysis environment
also includes the EMSE Suite comprising modules for integrating EEG with
structural MRI . A flowchart of the integrated analysis environment
is shown in Fig. 2. Dotted arrows indicate the points where structural
information and functional features derived from inverse-problem
computations will feed back into available forward-problem solvers.
Solutions to either the NIRS or EEG inverse problem
ideally would be based on knowledge of individualized boundary
conditions. To support instances where such information is not
available, we have implemented an alternative solution wherein a
selected atlas is substituted for individual-subject structural
information. The developed human-head atlas is summarized in Fig. 3. A
montage of standard EEG electrode positions is included (Fig. 3A) to
guide placement of NIRS optodes. A depiction of functional image data
interpolated onto the cortical surface, along with a montage of the
corresponding sensor array, is shown in Fig. 3B. The array information
also is displayed the on the atlas segment selected for image
reconstruction, as shown in Fig. 3C.
The full capabilities of the Testbed have been explored
by programming the phantom to support causally directed activities
involving three different head regions. Volume-averaged image time
series from the three colored regions in Fig. 4 were used as input for
DCM model-selection computations and showed that the preferred model was
 R. L. Barbour, R. Ansari, R. Al
abdi, H. L. Graber, M. B. Levin, Y. Pei, C. H. Schmitz, and Y. Xu,
“Validation of near infrared spectroscopic (NIRS) imaging using
programmable phantoms,” Paper 687002 in Design and Performance
Validation of Phantoms Used in Conjunction with Optical Measurements of
Tissue (Proceedings of SPIE, Vol. 6870), R.J. Nordstrom, Ed. (2008).
Y. Pei, Z. Wang, and R.L. Barbour, “NAVI: A problem solving environment
(PSE) for NIRS data analysis,” Poster No. 685 T-AM at Human Brain
Mapping 2006 (Florence, Italy, June 11-15, 2006).
 NIRx fNIRS Analysis Environment User’s guide. Available at: http://otg.downstate.edu/Publication/NIRxPackage_02.pdf.
W. D. Penny, K. J. Friston, J. T. Ashburner, S. J. Kiebel, and T. E.
Nichols, Statistical Parametric Mapping: The Analysis of Functional
Brain Images. Academic Press, 2006.
 N. Tzourio-Mazoyer, B.
Landeau, D. Papathanassiou, F. Crivello, O. Etard, N. Delcroix, B.
Mazoyer, and M. Joliot, “Automated anatomical labeling of activations in
SPM using a macroscopic anatomical parcellation of the MNI MRI
single-subject brain,” NeuroImage 15, 273-289 (2002).
 K. Friston,
“Causal modeling and brain connectivity in functional magnetic
resonance imaging,” PLoS Biology 7, 0220-0225 (2009).
Signal Imaging, Inc., EMSE®Suite User Manual, Version 5.4, San Diego,
2011. Available at: ftp://ftp.sourcesignal.com/manuals/