<s>
Positron	B-Application
emission	I-Application
tomography	I-Application
for	O
bone	O
imaging	O
,	O
as	O
an	O
in	O
vivo	O
tracer	O
technique	O
,	O
allows	O
the	O
measurement	O
of	O
the	O
regional	O
concentration	O
of	O
radioactivity	O
proportional	O
to	O
the	O
image	O
pixel	B-Algorithm
values	O
averaged	O
over	O
a	O
region	O
of	O
interest	O
(	O
ROI	O
)	O
in	O
bones	O
.	O
</s>
<s>
Positron	B-Application
emission	I-Application
tomography	I-Application
is	O
a	O
functional	B-Algorithm
imaging	I-Algorithm
technique	O
that	O
uses	O
[18F]NaF	O
radiotracer	O
to	O
visualise	O
and	O
quantify	O
regional	O
bone	O
metabolism	O
and	O
blood	O
flow	O
.	O
</s>
<s>
The	O
functional	B-Algorithm
imaging	I-Algorithm
technique	O
of	O
dynamic	O
[18F]NaF	O
PET	B-Application
scans	I-Application
can	O
quantify	O
regional	O
bone	O
turnover	O
at	O
specific	O
sites	O
of	O
clinical	O
importance	O
such	O
as	O
the	O
lumbar	O
spine	O
and	O
hip	O
and	O
has	O
been	O
validated	O
by	O
comparison	O
with	O
the	O
gold	O
standard	O
of	O
bone	O
biopsy	O
.	O
</s>
<s>
[18F]NaF	O
is	O
taken-up	O
in	O
an	O
exponential	O
manner	O
representing	O
the	O
equilibration	O
of	O
tracer	O
with	O
the	O
extracellular	O
and	O
cellular	O
fluid	O
spaces	O
with	O
a	O
half-life	B-Application
of	O
0.4	O
hours	O
,	O
and	O
with	O
kidneys	O
with	O
a	O
half-life	B-Application
of	O
2.4	O
hours	O
.	O
</s>
<s>
The	O
standardized	B-Algorithm
uptake	I-Algorithm
value	I-Algorithm
(	O
SUV	B-Algorithm
)	O
is	O
defined	O
as	O
tissue	O
concentration	O
(	O
KBq/ml	O
)	O
divided	O
by	O
activity	O
injected	O
normalized	O
for	O
body	O
weight	O
.	O
</s>
<s>
The	O
SUV	B-Algorithm
measured	O
from	O
the	O
large	O
ROI	O
smooths	O
out	O
the	O
noise	O
and	O
,	O
therefore	O
,	O
more	O
appropriate	O
in	O
[18F]NaF	O
bone	O
studies	O
as	O
the	O
radiotracer	O
is	O
fairly	O
uniformly	O
taken	O
up	O
throughout	O
the	O
bone	O
.	O
</s>
<s>
The	O
measurement	O
of	O
SUV	B-Algorithm
is	O
easy	O
,	O
cheap	O
,	O
and	O
quicker	O
to	O
perform	O
,	O
making	O
it	O
more	O
attractive	O
for	O
clinical	O
use	O
.	O
</s>
<s>
SUV	B-Algorithm
can	O
be	O
measured	O
at	O
a	O
single	O
site	O
,	O
or	O
the	O
whole	O
skeleton	O
using	O
a	O
series	O
of	O
static	O
scans	O
and	O
restricted	O
by	O
the	O
small	O
field-of-view	O
of	O
the	O
PET	B-Application
scanner	I-Application
.	O
</s>
<s>
The	O
SUV	B-Algorithm
has	O
emerged	O
as	O
a	O
clinically	O
useful	O
,	O
albeit	O
controversial	O
,	O
semi-quantitative	O
tool	O
in	O
PET	O
analysis	O
.	O
</s>
<s>
Standardizing	O
imaging	O
protocols	O
and	O
measuring	O
the	O
SUV	B-Algorithm
at	O
the	O
same	O
time	O
post-injection	O
of	O
the	O
radiotracer	O
,	O
is	O
necessary	O
to	O
obtain	O
a	O
correct	O
SUV	B-Algorithm
because	O
imaging	O
before	O
the	O
uptake	O
plateau	O
introduces	O
unpredictable	O
errors	O
of	O
up	O
to	O
50%	O
with	O
SUVs	B-Algorithm
.	O
</s>
<s>
Noise	O
,	O
image	O
resolution	O
,	O
and	O
reconstruction	O
do	O
affect	O
the	O
accuracy	O
of	O
SUVs	B-Algorithm
,	O
but	O
correction	O
with	O
phantom	O
can	O
minimize	O
these	O
differences	O
when	O
comparing	O
SUVs	B-Algorithm
for	O
multi-centre	O
clinical	O
trials	O
.	O
</s>
<s>
SUV	B-Algorithm
may	O
lack	O
sensitivity	O
in	O
measuring	O
response	O
to	O
treatment	O
as	O
it	O
is	O
a	O
simple	O
measure	O
of	O
tracer	O
uptake	O
in	O
bone	O
,	O
which	O
is	O
affected	O
by	O
the	O
tracer	O
uptake	O
in	O
other	O
competing	O
tissues	O
and	O
organs	O
in	O
addition	O
to	O
the	O
target	O
ROI	O
.	O
</s>
<s>
The	O
quantification	O
of	O
dynamic	O
PET	B-Application
studies	I-Application
to	O
measure	O
Ki	O
requires	O
the	O
measurement	O
of	O
the	O
skeletal	O
time-activity	B-Algorithm
curves	I-Algorithm
(	O
TAC	O
)	O
from	O
the	O
region	O
of	O
interest	O
(	O
ROI	O
)	O
and	O
the	O
arterial	B-Algorithm
input	I-Algorithm
function	I-Algorithm
(	O
AIF	O
)	O
,	O
which	O
can	O
be	O
measured	O
in	O
various	O
different	O
ways	O
.	O
</s>
<s>
However	O
,	O
the	O
most	O
common	O
is	O
to	O
correct	O
the	O
image-based	O
blood	O
time-activity	B-Algorithm
curves	I-Algorithm
using	O
several	O
venous	O
blood	O
samples	O
taken	O
at	O
discrete	O
time	O
points	O
while	O
the	O
patient	O
is	O
scanned	O
.	O
</s>
<s>
Measurement	O
of	O
the	O
arterial	B-Algorithm
input	I-Algorithm
function	I-Algorithm
(	O
AIF	O
)	O
,	O
which	O
acts	O
as	O
the	O
first	O
input	O
to	O
the	O
mathematical	O
model	O
of	O
tracer	O
distribution	O
.	O
</s>
<s>
Measurement	O
of	O
the	O
time-activity	B-Algorithm
curve	I-Algorithm
(	O
TAC	O
)	O
within	O
the	O
skeletal	O
region	O
of	O
interest	O
,	O
which	O
acts	O
as	O
the	O
second	O
input	O
to	O
the	O
mathematical	O
model	O
of	O
tracer	O
distribution	O
.	O
</s>
<s>
Since	O
A	O
tissue	O
TAC	O
can	O
be	O
expressed	O
as	O
a	O
convolution	O
of	O
measured	O
arterial	B-Algorithm
input	I-Algorithm
function	I-Algorithm
with	O
IRF	O
,	O
Cbone(t )	O
can	O
be	O
expressed	O
as	O
:	O
</s>
<s>
A	O
tissue	O
TAC	O
is	O
modelled	O
as	O
a	O
convolution	O
of	O
measured	O
arterial	B-Algorithm
input	I-Algorithm
function	I-Algorithm
with	O
IRF	O
,	O
the	O
estimates	O
for	O
IRF	O
are	O
obtained	O
iteratively	O
to	O
minimise	O
the	O
differences	O
between	O
the	O
left	O
-	O
and	O
right-hand	O
side	O
of	O
the	O
following	O
Equation	O
:	O
</s>
<s>
The	O
measurement	O
of	O
Ki	O
from	O
dynamic	O
PET	B-Application
scans	I-Application
require	O
tracer	O
kinetic	O
modelling	O
to	O
obtain	O
the	O
model	O
parameters	O
describing	O
the	O
biological	O
processes	O
in	O
bone	O
,	O
as	O
described	O
by	O
Hawkins	O
et	O
al	O
.	O
</s>
<s>
The	O
calculation	O
of	O
Ki	O
using	O
Patlak	O
method	O
is	O
simpler	O
than	O
using	O
non-linear	O
regression	O
(	O
NLR	O
)	O
fitting	O
the	O
arterial	B-Algorithm
input	I-Algorithm
function	I-Algorithm
and	O
the	O
tissue	O
time-activity	B-Algorithm
curve	I-Algorithm
data	O
to	O
the	O
Hawkins	O
model	O
.	O
</s>
<s>
The	O
calculation	O
of	O
Ki	O
using	O
arterial	B-Algorithm
input	I-Algorithm
function	I-Algorithm
,	O
time-activity	B-Algorithm
curve	I-Algorithm
,	O
and	O
Hawkins	O
model	O
was	O
limited	O
to	O
a	O
small	O
skeletal	O
region	O
covered	O
by	O
the	O
narrow	O
field-of-view	O
of	O
the	O
PET	B-Application
scanner	I-Application
while	O
acquiring	O
a	O
dynamic	O
scan	O
.	O
</s>
<s>
showed	O
in	O
2012	O
that	O
it	O
is	O
possible	O
to	O
measure	O
Ki	O
values	O
in	O
bones	O
using	O
static	O
[18F]NaF	O
PET	B-Application
scans	I-Application
.	O
</s>
<s>
The	O
Siddique	O
–	O
Blake	O
approach	O
is	O
based	O
on	O
the	O
combination	O
of	O
the	O
Patlak	O
method	O
,	O
the	O
semi-population	O
based	O
arterial	B-Algorithm
input	I-Algorithm
function	I-Algorithm
,	O
and	O
the	O
information	O
that	O
Vo	O
does	O
not	O
significantly	O
change	O
post-treatment	O
.	O
</s>
<s>
The	O
most	O
fundamental	O
difference	O
between	O
SUV	B-Algorithm
and	O
Ki	O
values	O
is	O
that	O
SUV	B-Algorithm
is	O
a	O
simple	O
measure	O
of	O
uptake	O
,	O
which	O
is	O
normalized	O
to	O
body	O
weight	O
and	O
injected	O
activity	O
.	O
</s>
<s>
The	O
SUV	B-Algorithm
does	O
not	O
take	O
into	O
consideration	O
the	O
tracer	O
delivery	O
to	O
the	O
local	O
region	O
of	O
interest	O
from	O
where	O
the	O
measurements	O
are	O
obtained	O
,	O
therefore	O
,	O
affected	O
by	O
the	O
physiological	O
process	O
consuming	O
[18F]NaF	O
elsewhere	O
in	O
the	O
body	O
.	O
</s>
<s>
The	O
difference	O
in	O
the	O
measurement	O
of	O
Ki	O
and	O
SUV	B-Algorithm
in	O
bone	O
tissue	O
using	O
[18F]NaF	O
are	O
explained	O
in	O
more	O
detail	O
by	O
Blake	O
et	O
al	O
.	O
</s>
<s>
However	O
,	O
the	O
calculation	O
of	O
SUV	B-Algorithm
requires	O
a	O
single	O
static	O
PET	B-Application
scan	I-Application
performed	O
approximately	O
45	O
–	O
60	O
minutes	O
post-tracer	O
injection	O
at	O
any	O
region	O
imaged	O
within	O
the	O
skeleton	O
.	O
</s>
<s>
Many	O
researchers	O
have	O
shown	O
a	O
high	O
correlation	O
between	O
SUV	B-Algorithm
and	O
Ki	O
values	O
at	O
various	O
skeletal	O
sites	O
.	O
</s>
<s>
However	O
,	O
SUV	B-Algorithm
and	O
Ki	O
methods	O
can	O
contradict	O
for	O
measuring	O
response	O
to	O
treatment	O
.	O
</s>
<s>
Since	O
SUV	B-Algorithm
has	O
not	O
been	O
validated	O
against	O
the	O
histomorphometry	O
,	O
its	O
usefulness	O
in	O
bone	O
studies	O
measuring	O
response	O
to	O
treatment	O
and	O
disease	O
progression	O
is	O
uncertain	O
.	O
</s>
