Decoding of movement direction using optical imaging of motor cortex Page: 1

                                
Tam and Zouridakis BMC Neuroscience 2013, 14(Suppl 1):P380
http://www.biomedcentral.com/1471-2202/14/S1 /P380

BNeuMC
Neuroscience

"r, ] il :, :  I- : 1  1 / I //I  -                                       I ~LI
Decoding of movement direction using optical
imaging of motor cortex
Nicoladie D Tam', George Zouridakis2
From Twenty Second Annual Computational Neuroscience Meeting: CNS*2013
Paris, France. 13-18 July 2013

This study provides a computational scheme to decode
intentional arm movement direction using optical imaging
of the motor cortex for future implementation on a neuro-
prosthetic device that enables physically disabled patients
to navigate a wheelchair using brain-derived signals. To
this end, we developed a signal-processing algorithm for
detecting movement direction from hemodynamic signals
using functional near-infrared spectroscopy (fNIRS)
recorded in human subjects during execution of a direc-
tional motor task. fNIRS has been shown to reflect hemo-
dynamic responses in the cortex during execution of
computational [1,4,6] and motor tasks [2], and can capture
better regional cerebral blood volume (rCBV) changes [3]
than blood-oxygen level dependent (BOLD) signals [5].
We used 64 spatially distributed optrodes to record from
both hemispheres of motor cortices during free arm
orthogonal movements in the x and y directions on a hori-
zontal plane, extending the experimental findings reported
earlier [7]. We then analyzed the spatiotemporal profiles
of the 64-channel hemodynamic response to derive the
direction of the movement executed from the motor cor-
tex activation.
We employed four different measures of hemodynamic
profiles - oxy- (HbO2) and deoxy-hemoglobin (Hb), and
their sum (HbO2 + Hb) and difference (HbO2 - Hb) sig-
nals - to correlate oxygen delivery, oxygen extraction,
total blood volume delivered, and total oxygenation with
a series of specific movements to identify the direction of
the intentional movements. This analysis provided a
unique representation of the different hemodynamic
components of the localized neuronal populations in the
motor cortices underlying the optrodes.

Our results show that the four measures of hemody-
namic response may be coupled in one movement direc-
tion and decoupled in another, for the same subject. That
is, oxygen delivery, extraction, total blood volume, and
oxygenation do not necessarily co-vary (increase or
decrease simultaneously) for all movement directions,
but for some directions, they may be decoupled; e.g.,
oxygen delivery may increase while, at the same time,
oxygen extraction may decrease. This suggested that oxy-
gen extraction may outpace oxygen delivery due to high
oxygen demand of the underlying neural tissues, resulting
in decoupling of the oxygen delivery and extraction vari-
ables. Thus, if decoding of movement direction relies on
a single hemodynamic measure, the latter is not sufficient
to identify the movement direction uniquely. Instead, a
combination of all four measures of hemodynamic signals
is needed, and it can be extracted from the temporal pro-
files of neural activation and deactivation that represent
temporal coupling and decoupling of oxygen delivery and
extraction. Overall, our experiments demonstrate the fea-
sibility of decoding intentional movement direction using
computational analysis of optical recordings from the
motor cortex, instead of implanting microelectrodes, and
support a future implementation of a hands-free neuro-
prosthetic device for the physically disabled to navigate a
wheelchair unassisted.
Author details
Department of Biological Sciences, University of North Texas, Denton, TX
76203, USA. 2Departments of Engineering Technology, Computer Science,
and Electrical and Computer Engineering, University of Houston, Houston,
TX, 77204, USA.

Published: 8 July 2013
* Correspondence: nicoladie.tam@unt.edu
1Department of Biological Sciences, University of North Texas, Denton, TX
76203, USA
Full list of author information is available at the end of the article
 2013 Tam and Zouridakis; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
BioM      ed   Central       Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.