<s>
Seed-based	B-Algorithm
d	I-Algorithm
mapping	I-Algorithm
(	O
formerly	O
Signed	B-Algorithm
differential	I-Algorithm
mapping	I-Algorithm
)	O
or	O
SDM	O
is	O
a	O
statistical	O
technique	O
created	O
by	O
Joaquim	O
Radua	O
for	O
meta-analyzing	O
studies	O
on	O
differences	O
in	O
brain	O
activity	O
or	O
structure	O
which	O
used	O
neuroimaging	B-Algorithm
techniques	O
such	O
as	O
fMRI	B-Algorithm
,	O
VBM	B-Algorithm
,	O
DTI	O
or	O
PET	B-Application
.	O
</s>
<s>
One	O
of	O
the	O
new	O
features	O
,	O
introduced	O
to	O
avoid	O
positive	O
and	O
negative	O
findings	O
in	O
the	O
same	O
voxel	B-Algorithm
as	O
seen	O
in	O
previous	O
methods	O
,	O
was	O
the	O
representation	O
of	O
both	O
positive	O
differences	O
and	O
negative	O
differences	O
in	O
the	O
same	O
map	O
,	O
thus	O
obtaining	O
a	O
signed	O
differential	O
map	O
(	O
SDM	O
)	O
.	O
</s>
<s>
Another	O
relevant	O
feature	O
,	O
introduced	O
in	O
version	O
2.11	O
,	O
was	O
the	O
use	O
of	O
effect	B-Application
sizes	I-Application
(	O
leading	O
to	O
effect-size	O
SDM	O
or	O
'	O
ES-SDM	O
'	O
)	O
,	O
which	O
allows	O
combination	O
of	O
reported	O
peak	O
coordinates	O
with	O
statistical	O
parametric	O
maps	O
,	O
thus	O
allowing	O
more	O
exhaustive	O
and	O
accurate	O
meta-analyses	O
.	O
</s>
<s>
the	O
voxels	B-Algorithm
where	O
the	O
differences	O
between	O
patients	O
and	O
healthy	O
controls	O
were	O
highest	O
)	O
,	O
and	O
statistical	O
maps	O
if	O
available	O
,	O
are	O
selected	O
according	O
to	O
SDM	O
inclusion	O
criteria	O
.	O
</s>
<s>
Second	O
,	O
coordinates	O
are	O
used	O
to	O
recreate	O
statistical	O
maps	O
,	O
and	O
effect-sizes	O
maps	O
and	O
their	O
variances	O
are	O
derived	O
from	O
t-statistics	O
(	O
or	O
equivalently	O
from	O
p-values	B-General_Concept
or	O
z-scores	O
)	O
.	O
</s>
<s>
Finally	O
,	O
individual	O
study	O
maps	O
are	O
meta-analyzed	O
using	O
different	O
tests	O
to	O
complement	O
the	O
main	O
outcome	O
with	O
sensitivity	O
and	O
heterogeneity	B-General_Concept
analyses	O
.	O
</s>
<s>
It	O
is	O
not	O
uncommon	O
in	O
neuroimaging	B-Algorithm
studies	O
that	O
some	O
regions	O
(	O
e.g.	O
</s>
<s>
a	O
priori	O
regions	B-Algorithm
of	I-Algorithm
interest	I-Algorithm
)	O
are	O
more	O
liberally	O
thresholded	O
than	O
the	O
rest	O
of	O
the	O
brain	O
.	O
</s>
<s>
Pre-processing	O
of	O
statistical	O
parametric	O
maps	O
is	O
straightforward	O
,	O
while	O
pre-processing	O
of	O
reported	O
peak	O
coordinates	O
requires	O
recreating	O
the	O
clusters	O
of	O
difference	O
by	O
means	O
of	O
an	O
un-normalized	O
Gaussian	O
Kernel	O
,	O
so	O
that	O
voxels	B-Algorithm
closer	O
to	O
the	O
peak	O
coordinate	O
have	O
higher	O
values	O
.	O
</s>
<s>
coregistration	B-Algorithm
mismatch	O
in	O
the	O
studies	O
,	O
the	O
size	O
of	O
the	O
cluster	O
or	O
the	O
location	O
of	O
the	O
peak	O
within	O
the	O
cluster	O
.	O
</s>
<s>
Within	O
a	O
study	O
,	O
values	O
obtained	O
by	O
close	O
Gaussian	O
kernels	O
are	O
summed	O
,	O
though	O
values	O
are	O
combined	O
by	O
square-distance-weighted	O
averaging	O
.	O
</s>
<s>
SDM	O
provides	O
several	O
different	O
statistical	O
analyses	O
in	O
order	O
to	O
complement	O
the	O
main	O
outcome	O
with	O
sensitivity	O
and	O
heterogeneity	B-General_Concept
analyses	O
.	O
</s>
<s>
The	O
main	O
statistical	O
analysis	O
is	O
the	O
mean	O
analysis	O
,	O
which	O
consists	O
in	O
calculating	O
the	O
mean	O
of	O
the	O
voxel	B-Algorithm
values	O
in	O
the	O
different	O
studies	O
.	O
</s>
<s>
This	O
mean	O
is	O
weighted	B-Algorithm
by	O
the	O
inverse	O
of	O
the	O
variance	O
and	O
accounts	O
for	O
inter-study	B-General_Concept
heterogeneity	I-General_Concept
(	O
QH	O
maps	O
)	O
.	O
</s>
<s>
Subgroup	O
analyses	O
are	O
mean	O
analyses	O
applied	O
to	O
groups	O
of	O
studies	O
to	O
allow	O
the	O
study	O
of	O
heterogeneity	B-General_Concept
.	O
</s>
<s>
Linear	B-Algorithm
model	I-Algorithm
analyses	O
(	O
e.g.	O
</s>
<s>
The	O
statistical	B-General_Concept
significance	I-General_Concept
of	O
the	O
analyses	O
is	O
checked	O
by	O
standard	O
randomization	B-Algorithm
tests	I-Algorithm
.	O
</s>
<s>
It	O
is	O
recommended	O
to	O
use	O
uncorrected	O
p-values	B-General_Concept
=	O
0.005	O
,	O
as	O
this	O
significance	O
has	O
been	O
found	O
in	O
this	O
method	O
to	O
be	O
approximately	O
equivalent	O
to	O
a	O
corrected	O
p-value	B-General_Concept
=	O
0.05	O
.	O
</s>
<s>
A	O
false	B-General_Concept
discovery	I-General_Concept
rate	I-General_Concept
(	O
FDR	O
)	O
=	O
0.05	O
has	O
been	O
found	O
in	O
this	O
method	O
to	O
be	O
too	O
conservative	O
.	O
</s>
<s>
SDM	O
is	O
software	O
written	O
by	O
the	O
SDM	O
project	O
to	O
aid	O
the	O
meta-analysis	O
of	O
voxel-based	B-Algorithm
neuroimaging	B-Algorithm
data	O
.	O
</s>
<s>
It	O
is	O
distributed	O
as	O
freeware	B-License
including	O
a	O
graphical	O
interface	O
and	O
a	O
menu/command	O
-line	O
console	O
.	O
</s>
