[BIC-announce] FW: Lecture - Estimating Effective Connectivity Over Brain Manifolds

Jennifer Chew, Ms. jennifer.chew@mcgill.ca
Thu, 16 Jun 2005 16:12:07 -0400


This is a multi-part message in MIME format.

------_=_NextPart_001_01C572AF.AC0FAC1C
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

 ***LECTURE***     =20
=20
Date:  Friday, June 17, 2005
Time:  10:00 A.M.
Place:  Webster 200
=20
=20
Speaker:  Pedro A. Valdes-Sosa, Vice-Director, Cuban Neuroscience Centre
=20
 Title:   Estimating Effective Connectivity over Brain Manifolds

=20

There is much current interest in identifying the anatomical and =
functional circuits that are the basis of the brain's computations, with =
hope that functional Neuroimaging techniques will allow the in vivo =
study of these neural processes through the statistical analysis of the =
time series they produce. Ideally, the use of techniques such as =
Multivariate Autoregressive Modeling (MAR) should allow the =
identification of effective connectivity by combining Graphical modeling =
methods with the concept of Granger Causality. Unfortunately, current =
time series methods perform well only for the case that the length of =
the time series Nt is much larger than p, the number of brain sites =
studied, which is exactly the reverse of the situation in Neuroimaging =
for which relatively short time series are measured over thousands of =
voxels. Usual methods also ignore the full spatio-temporal nature of =
functional brain data which are, in fact, collections of time series =
sampled over an underlying continuous spatial manifold-the brain. A =
fully spatio-temporal MAR model (ST-MAR) is developed within the =
framework of functional data analysis. For spatial data, each row of a =
matrix Ak is the influence field of a given voxel. A Bayesian ST-MAR =
model is specified in which the influence fields for all voxels are =
required to vary smoothly over space. This requirement is enforced by =
penalizing the spatial roughness of the influence fields. This roughness =
is calculated with a discrete version of the spatial Laplacian operator. =
Additional constraints are also introduced by recognizing the fact that =
neural connections are comparatively sparse, and introducing the class =
of Sparse Multivariate Autoregressive models. These can be estimated in =
a two stage process involving a) penalized regression and b) pruning of =
unlikely connections by means of the local false discovery rate =
developed by Efron. Extensive simulations were performed with idealized =
cortical networks having small world topologies and stable dynamics. =
These show that the detection efficiency of connections of the proposed =
procedure is quite high. Application of the method to real data was =
illustrated by the identification of neural circuitry related to a) =
emotional processing as measured by BOLD and b) the origin of EEG =
rhythms obtained during concurrent EEG/fMRI recordings.

Refer=EAncias

[1]                    P.A: Valdes-Sosa, PA, (2004), Spatio-temporal =
autoregressive models defined over brain manifolds: Neuroinformatics, v. =
2, p. 239-250.

.[2]                   P.A. Valdes-Sosa, J.M. Sanchez-Bornot, A. =
Lage-Castellanos, M. Vega-Hernandez, J. Bosch-Bayard, L. Melie-Garc=EDa =
and E. Canales-Rodriguez.(2005) "Estimating Brain Functional =
Connectivity with Sparse Multivariate Autoregression", Philosophical =
Transactions of the Royal Society B. Theme Issue on Multimodal Brain =
Connectivity. (Eds) P. Valdes-Sosa, R. Kotter, K. Friston, in press

=20

=20

=20

=20

=20


------_=_NextPart_001_01C572AF.AC0FAC1C
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML xmlns=3D"http://www.w3.org/TR/REC-html40" xmlns:o =3D=20
"urn:schemas-microsoft-com:office:office" xmlns:w =3D=20
"urn:schemas-microsoft-com:office:word" xmlns:st1 =3D=20
"urn:schemas-microsoft-com:office:smarttags" xmlns:ns0 =3D=20
"urn:schemas-microsoft-com:office:smarttags"><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 6.00.2800.1498" name=3DGENERATOR><o:SmartTagType =

namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"=20
name=3D"State"></o:SmartTagType><o:SmartTagType=20
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"=20
name=3D"Street"></o:SmartTagType><o:SmartTagType=20
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"=20
name=3D"PlaceType"></o:SmartTagType><o:SmartTagType=20
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"=20
name=3D"PlaceName"></o:SmartTagType><o:SmartTagType=20
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"=20
name=3D"country-region"></o:SmartTagType><o:SmartTagType=20
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"=20
name=3D"City"></o:SmartTagType><o:SmartTagType=20
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"=20
name=3D"place"></o:SmartTagType><o:SmartTagType=20
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"=20
name=3D"address"></o:SmartTagType><!--[if !mso]>
<STYLE>st1\:* {
	BEHAVIOR: url(#default#ieooui)
}
</STYLE>
<![endif]-->
<STYLE>@font-face {
	font-family: Times;
}
@font-face {
	font-family: Arial Narrow;
}
@font-face {
	font-family: Palatino-Roman;
}
@page Section1 {size: 612.0pt 792.0pt; margin: 72.0pt 90.0pt 72.0pt =
90.0pt; }
P.MsoNormal {
	FONT-SIZE: 12pt; MARGIN: 6pt 0pt; FONT-FAMILY: Arial
}
LI.MsoNormal {
	FONT-SIZE: 12pt; MARGIN: 6pt 0pt; FONT-FAMILY: Arial
}
DIV.MsoNormal {
	FONT-SIZE: 12pt; MARGIN: 6pt 0pt; FONT-FAMILY: Arial
}
H1 {
	FONT-WEIGHT: bold; FONT-SIZE: 20pt; MARGIN: 6pt 0pt 12pt; FONT-FAMILY: =
Arial; TEXT-ALIGN: center
}
H2 {
	MARGIN-TOP: 6pt; FONT-WEIGHT: bold; FONT-SIZE: 18pt; MARGIN-LEFT: 36pt; =
TEXT-INDENT: -36pt; MARGIN-RIGHT: 0pt; FONT-FAMILY: Arial; =
mso-margin-bottom-alt: auto
}
H3 {
	FONT-WEIGHT: bold; FONT-SIZE: 14pt; MARGIN: 6pt 0pt; FONT-STYLE: =
italic; FONT-FAMILY: Arial
}
H4 {
	FONT-WEIGHT: bold; FONT-SIZE: 14pt; MARGIN: 12pt 0pt 3pt; FONT-FAMILY: =
Arial
}
P.MsoHeader {
	FONT-WEIGHT: bold; FONT-SIZE: 14pt; MARGIN: 6pt 0pt; FONT-FAMILY: Arial
}
LI.MsoHeader {
	FONT-WEIGHT: bold; FONT-SIZE: 14pt; MARGIN: 6pt 0pt; FONT-FAMILY: Arial
}
DIV.MsoHeader {
	FONT-WEIGHT: bold; FONT-SIZE: 14pt; MARGIN: 6pt 0pt; FONT-FAMILY: Arial
}
A:link {
	COLOR: blue; TEXT-DECORATION: underline
}
SPAN.MsoHyperlink {
	COLOR: blue; TEXT-DECORATION: underline
}
A:visited {
	COLOR: purple; TEXT-DECORATION: underline
}
SPAN.MsoHyperlinkFollowed {
	COLOR: purple; TEXT-DECORATION: underline
}
P {
	FONT-SIZE: 12pt; MARGIN-LEFT: 0pt; MARGIN-RIGHT: 0pt; FONT-FAMILY: =
"Times New Roman"; mso-margin-bottom-alt: auto; mso-margin-top-alt: auto
}
P.StyleHeading1Centered {
	FONT-WEIGHT: bold; FONT-SIZE: 18pt; MARGIN: 0pt; FONT-FAMILY: Arial; =
TEXT-ALIGN: center
}
LI.StyleHeading1Centered {
	FONT-WEIGHT: bold; FONT-SIZE: 18pt; MARGIN: 0pt; FONT-FAMILY: Arial; =
TEXT-ALIGN: center
}
DIV.StyleHeading1Centered {
	FONT-WEIGHT: bold; FONT-SIZE: 18pt; MARGIN: 0pt; FONT-FAMILY: Arial; =
TEXT-ALIGN: center
}
SPAN.EmailStyle19 {
	COLOR: windowtext; FONT-FAMILY: Arial; mso-style-type: personal-compose
}
DIV.Section1 {
	page: Section1
}
</STYLE>
</HEAD>
<BODY lang=3DEN-US vLink=3Dpurple link=3Dblue>
<DIV dir=3Dltr align=3Dleft><SPAN style=3D"FONT-SIZE: 10pt"><SPAN=20
class=3D377575014-16062005><FONT face=3DArial><FONT =
color=3D#0000ff><SPAN=20
class=3D125560720-16062005>&nbsp;***LECTURE***&nbsp;</SPAN>&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp;</FONT></FONT></SPAN></SPAN></DIV>
<DIV><FONT face=3DArial color=3D#0000ff><SPAN style=3D"FONT-SIZE: =
10pt"><SPAN=20
class=3D377575014-16062005></SPAN></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial color=3D#0000ff><SPAN style=3D"FONT-SIZE: =
10pt"><SPAN=20
class=3D377575014-16062005><SPAN class=3D125560720-16062005>Date:&nbsp; =
Friday, June=20
17, 2005</SPAN></SPAN></SPAN></FONT></DIV>
<DIV><FONT face=3DArial color=3D#0000ff><SPAN style=3D"FONT-SIZE: =
10pt"><SPAN=20
class=3D377575014-16062005><SPAN class=3D125560720-16062005>Time:&nbsp; =
10:00=20
A.M.</SPAN></SPAN></SPAN></FONT></DIV>
<DIV><FONT face=3DArial color=3D#0000ff><SPAN style=3D"FONT-SIZE: =
10pt"><SPAN=20
class=3D377575014-16062005><SPAN class=3D125560720-16062005>Place:&nbsp; =
Webster=20
200</SPAN></SPAN></SPAN></FONT></DIV>
<DIV><FONT face=3DArial color=3D#0000ff><SPAN style=3D"FONT-SIZE: =
10pt"><SPAN=20
class=3D377575014-16062005></SPAN></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial color=3D#0000ff><SPAN style=3D"FONT-SIZE: =
10pt"><SPAN=20
class=3D377575014-16062005></SPAN></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial color=3D#0000ff><SPAN style=3D"FONT-SIZE: =
10pt"><SPAN=20
class=3D377575014-16062005><SPAN =
class=3D125560720-16062005>Speaker:&nbsp; Pedro A.=20
Valdes-Sosa, Vice-Director, Cuban Neuroscience=20
Centre</SPAN></SPAN></SPAN></FONT></DIV>
<DIV><FONT face=3DArial color=3D#0000ff><SPAN style=3D"FONT-SIZE: =
10pt"><SPAN=20
class=3D377575014-16062005><SPAN=20
class=3D125560720-16062005></SPAN></SPAN></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT size=3D3><SPAN style=3D"FONT-SIZE: 12pt"><FONT =
face=3DArial><SPAN=20
class=3D125560720-16062005><FONT color=3D#0000ff =
size=3D2>&nbsp;Title:&nbsp;=20
&nbsp;</FONT></SPAN>Estimating Effective Connectivity over Brain=20
Manifolds</FONT></SPAN></FONT><B><FONT size=3D2><SPAN lang=3DPT=20
style=3D"FONT-WEIGHT: bold; FONT-SIZE: =
10pt"><o:p></o:p></SPAN></FONT></B></DIV>
<DIV class=3DSection1>
<P class=3DMsoNormal style=3D"TEXT-ALIGN: center" align=3Dcenter>
<P class=3DMsoNormal style=3D"TEXT-ALIGN: center" =
align=3Dcenter><I><FONT face=3DArial=20
size=3D2><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-STYLE: =
italic"><o:p></o:p></SPAN></FONT></I></P><FONT=20
face=3DArial size=3D2><SPAN lang=3DPT=20
style=3D"FONT-SIZE: 10pt"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D3><SPAN =
style=3D"FONT-SIZE: 12pt">There=20
is much current interest in identifying the anatomical and functional =
circuits=20
that are the basis of the brain=92s computations, with hope that =
functional=20
Neuroimaging techniques will allow the in vivo study of these neural =
processes=20
through the statistical analysis of the time series they produce. =
Ideally, the=20
use of techniques such as Multivariate Autoregressive Modeling (MAR) =
should=20
allow the identification of effective connectivity by combining =
Graphical=20
modeling methods with the concept of Granger Causality. Unfortunately, =
current=20
time series methods perform well only for the case that the length of =
the time=20
series Nt is much larger than p, the number of brain sites studied, =
which is=20
exactly the reverse of the situation in Neuroimaging for which =
relatively short=20
time series are measured over thousands of voxels. </SPAN>Usual methods =
also=20
ignore the full spatio-temporal nature of functional brain data which =
are, in=20
fact, collections of time series sampled over an underlying continuous =
spatial=20
manifold=97the brain. A fully spatio-temporal MAR model (ST-MAR) is =
developed=20
within the framework of functional data analysis. For spatial data, each =
row of=20
a matrix <B><SPAN style=3D"FONT-WEIGHT: bold">A</SPAN></B><I><SPAN=20
style=3D"FONT-STYLE: italic">k </SPAN></I>is the <I><SPAN=20
style=3D"FONT-STYLE: italic">influence field </SPAN></I>of a given =
voxel.=20
<st1:Street w:st=3D"on"><st1:address w:st=3D"on">A Bayesian=20
ST</st1:address></st1:Street>-MAR model is specified in which the =
influence=20
fields for all voxels are required to vary smoothly over space. This =
requirement=20
is enforced by penalizing the spatial roughness of the influence fields. =
This=20
roughness is calculated with a discrete version of the spatial Laplacian =

operator. Additional constraints </FONT>are also introduced by =
recognizing the=20
fact that neural connections are comparatively sparse, and introducing =
the class=20
of Sparse Multivariate Autoregressive models. These can be estimated in =
a two=20
stage process involving a) penalized regression and b) pruning of =
unlikely=20
connections by means of the local false discovery rate developed by =
Efron.=20
Extensive simulations were performed with idealized cortical networks =
having=20
small world topologies and stable dynamics. These show that the =
detection=20
efficiency of connections of the proposed procedure is quite high. =
Application=20
of the method to real data was illustrated by the identification of =
neural=20
circuitry related to a) emotional processing as measured by BOLD and b) =
the=20
origin of EEG rhythms obtained during concurrent EEG/fMRI=20
recordings.<o:p></o:p></P>
<P style=3D"TEXT-ALIGN: justify"><FONT face=3DArial size=3D2><SPAN=20
style=3D"FONT-SIZE: 11pt; FONT-FAMILY: =
Arial">Refer=EAncias<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal style=3D"MARGIN-BOTTOM: 12pt"><FONT face=3DArial =
size=3D2><SPAN=20
style=3D"FONT-SIZE: 11pt">[1]&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; P.A:=20
</SPAN></FONT><st1:place w:st=3D"on"><st1:City =
w:st=3D"on">Valdes-Sosa</st1:City>,=20
<st1:State w:st=3D"on">PA</st1:State></st1:place>, (2004), =
Spatio-temporal=20
autoregressive models defined over brain manifolds: <I><SPAN=20
style=3D"FONT-STYLE: italic">Neuroinformatics</SPAN></I>, v. <B><SPAN=20
style=3D"FONT-WEIGHT: bold">2</SPAN></B>, p. 239-250.<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN-BOTTOM: 12pt"><FONT face=3DArial =
size=3D2><SPAN=20
style=3D"FONT-SIZE: =
11pt">.[2]&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
</SPAN></FONT>P.A. Valdes-Sosa, J.M. Sanchez-Bornot, A. =
Lage-Castellanos, M.=20
Vega-Hernandez, J. Bosch-Bayard, L. Melie-Garc=EDa and E. =
Canales-Rodriguez.(2005)=20
=93Estimating Brain Functional Connectivity with Sparse Multivariate=20
Autoregression=94, Philosophical Transactions of the Royal Society B. =
Theme Issue=20
on Multimodal Brain Connectivity. (Eds) P. Valdes-Sosa, R. Kotter, K. =
Friston,=20
in press<FONT size=3D2><SPAN style=3D"FONT-SIZE: =
11pt"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN=20
style=3D"FONT-SIZE: 10pt"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN=20
style=3D"FONT-SIZE: 10pt"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN=20
style=3D"FONT-SIZE: 10pt"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P><FONT face=3D"Times New Roman" color=3Dblack size=3D3><SPAN=20
style=3D"FONT-SIZE: 12pt; COLOR: =
black"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D3><SPAN=20
style=3D"FONT-SIZE: =
12pt"><o:p>&nbsp;</o:p></SPAN></FONT></P></DIV></BODY></HTML>

------_=_NextPart_001_01C572AF.AC0FAC1C--