<s>
In	O
the	O
field	O
of	O
medicine	O
,	O
radiomics	B-Algorithm
is	O
a	O
method	O
that	O
extracts	O
a	O
large	O
number	O
of	O
features	O
from	O
medical	B-Application
images	I-Application
using	O
data-characterisation	O
algorithms	O
.	O
</s>
<s>
The	O
hypothesis	O
of	O
radiomics	B-Algorithm
is	O
that	O
the	O
distinctive	O
imaging	O
features	O
between	O
disease	O
forms	O
may	O
be	O
useful	O
for	O
predicting	O
prognosis	O
and	O
therapeutic	O
response	O
for	O
various	O
cancer	O
types	O
,	O
thus	O
providing	O
valuable	O
information	O
for	O
personalized	O
therapy	O
.	O
</s>
<s>
Radiomics	B-Algorithm
emerged	O
from	O
the	O
medical	O
fields	O
of	O
radiology	O
and	O
oncology	O
and	O
is	O
the	O
most	O
advanced	O
in	O
applications	O
within	O
these	O
fields	O
.	O
</s>
<s>
Because	O
of	O
the	O
large	O
image	O
data	O
that	O
needs	O
to	O
be	O
processed	O
,	O
it	O
would	O
be	O
too	O
much	O
work	O
to	O
perform	O
the	O
segmentation	O
manually	O
for	O
every	O
single	O
image	O
if	O
a	O
radiomics	B-Algorithm
database	O
with	O
lots	O
of	O
data	O
is	O
created	O
.	O
</s>
<s>
The	O
results	O
should	O
be	O
generated	O
as	O
fast	O
as	O
possible	O
so	O
that	O
the	O
whole	O
process	O
of	O
radiomics	B-Algorithm
can	O
also	O
be	O
accelerated	O
.	O
</s>
<s>
After	O
the	O
segmentation	O
,	O
many	O
features	O
can	O
be	O
extracted	O
and	O
the	O
relative	O
net	O
change	O
from	O
longitudinal	O
images	O
(	O
delta-radiomics	O
)	O
can	O
be	O
computed	O
.	O
</s>
<s>
gray-level	B-Algorithm
co-occurrence	I-Algorithm
matrix	I-Algorithm
(	O
GLCM	O
)	O
,	O
run	O
length	O
matrix	O
(	O
RLM	O
)	O
,	O
size	O
zone	O
matrix	O
(	O
SZM	O
)	O
,	O
and	O
neighborhood	O
gray	O
tone	O
difference	O
matrix	O
(	O
NGTDM	O
)	O
derived	O
textures	O
,	O
textures	O
extracted	O
from	O
filtered	O
images	O
,	O
and	O
fractal	O
features	O
.	O
</s>
<s>
A	O
detailed	O
description	O
of	O
texture	O
features	O
for	O
radiomics	B-Algorithm
can	O
be	O
found	O
in	O
Parekh	O
et	O
al	O
.	O
</s>
<s>
Several	O
steps	O
are	O
necessary	O
to	O
create	O
an	O
integrated	O
radiomics	B-Algorithm
database	O
.	O
</s>
<s>
The	O
goal	O
of	O
radiomics	B-Algorithm
is	O
to	O
be	O
able	O
to	O
use	O
this	O
database	O
for	O
new	O
patients	O
.	O
</s>
<s>
They	O
also	O
confirmed	O
that	O
the	O
prognostic	O
ability	O
of	O
these	O
radiomics	B-Algorithm
features	O
may	O
be	O
transferred	O
from	O
lung	O
to	O
head-and-neck	O
cancer	O
.	O
</s>
<s>
Several	O
studies	O
have	O
also	O
showed	O
that	O
radiomic	O
features	O
are	O
better	O
at	O
predicting	O
treatment	O
response	O
than	O
conventional	O
measures	O
,	O
such	O
as	O
tumor	O
volume	O
and	O
diameter	O
,	O
and	O
the	O
maximum	O
radiotracer	O
uptake	O
on	O
positron	B-Application
emission	I-Application
tomography	I-Application
(	O
PET	O
)	O
imaging	O
.	O
</s>
<s>
Using	O
this	O
technique	O
an	O
algorithm	O
has	O
been	O
developed	O
,	O
after	O
initial	O
training	O
based	O
on	O
intra	O
tumor	O
lymphocyte	O
density	O
,	O
to	O
predict	O
the	O
probability	O
of	O
tumor	O
response	O
to	O
immunotherapy	O
,	O
providing	O
a	O
demonstration	O
of	O
the	O
clinical	O
potential	O
of	O
radiomics	B-Algorithm
as	O
a	O
powerful	O
tool	O
for	O
personalized	O
therapy	O
in	O
the	O
emerging	O
field	O
of	O
immunooncology	O
.	O
</s>
<s>
Other	O
studies	O
have	O
also	O
demonstrated	O
the	O
utility	O
of	O
radiomics	B-Algorithm
for	O
predicting	O
immunotherapy	O
response	O
of	O
NSCLC	O
patients	O
using	O
pre-treatment	O
CT	O
and	O
PET/CT	O
images	O
.	O
</s>
<s>
Radiomics	B-Algorithm
remains	O
inferior	O
to	O
conventional	O
techniques	O
in	O
some	O
applications	O
,	O
suggesting	O
the	O
necessity	O
of	O
continued	O
improvement	O
and	O
manipulation	O
of	O
Radiomics	B-Algorithm
features	O
to	O
different	O
clinical	O
scenarios	O
.	O
</s>
<s>
(	O
2020	O
)	O
demonstrated	O
that	O
morphological	O
Radiomics	B-Algorithm
features	O
were	O
inferior	O
to	O
previously	O
established	O
features	O
in	O
the	O
discrimination	O
of	O
intracranial	O
aneurysm	O
rupture	O
status	O
from	O
3-dimensional	O
rotational	B-Algorithm
angiography	I-Algorithm
.	O
</s>
<s>
Radiomics	B-Algorithm
offers	O
the	O
advantage	O
to	O
be	O
non	O
invasive	O
and	O
can	O
therefore	O
be	O
repeated	O
prospectively	O
for	O
a	O
given	O
patient	O
more	O
easily	O
than	O
invasive	O
tumor	O
biopsies	O
.	O
</s>
<s>
It	O
has	O
been	O
suggested	O
that	O
radiomics	B-Algorithm
could	O
be	O
a	O
mean	O
to	O
monitor	O
tumor	O
dynamic	O
changes	O
along	O
the	O
course	O
of	O
radiotherapy	O
and	O
to	O
define	O
sub	O
volumes	O
at	O
risk	O
for	O
which	O
dose	O
escalation	O
could	O
be	O
beneficial	O
.	O
</s>
<s>
Radiomics	B-Algorithm
demonstrated	O
significant	O
differences	O
in	O
a	O
set	O
of	O
82	O
treated	O
lesions	O
in	O
66	O
patients	O
with	O
pathological	O
outcomes	O
.	O
</s>
<s>
These	O
results	O
show	O
that	O
radiomics	B-Algorithm
holds	O
promise	O
for	O
differentiating	O
between	O
treatment	O
effect	O
and	O
true	O
progression	O
in	O
brain	O
metastases	O
treated	O
with	O
SRS	O
.	O
</s>
<s>
Radiomics	B-Algorithm
can	O
also	O
be	O
used	O
to	O
identify	O
challenging	O
physiological	O
events	O
such	O
as	O
brain	O
activity	O
,	O
which	O
is	O
usually	O
studied	O
with	O
imaging	O
techniques	O
such	O
as	O
functional	O
MRI	O
"	O
fMRI	O
"	O
.	O
</s>
<s>
Multiparametric	O
radiological	B-Application
imaging	I-Application
is	O
vital	O
for	O
detection	O
,	O
characterization	O
and	O
diagnosis	O
of	O
many	O
different	O
diseases	O
.	O
</s>
<s>
However	O
,	O
current	O
methods	O
in	O
radiomics	B-Algorithm
are	O
limited	O
to	O
using	O
single	O
images	O
for	O
the	O
extraction	O
of	O
these	O
textural	O
features	O
and	O
may	O
limit	O
the	O
applicable	O
scope	O
of	O
radiomics	B-Algorithm
in	O
different	O
clinical	O
settings	O
.	O
</s>
<s>
The	O
Multiparametric	O
Radiomics	B-Algorithm
was	O
tested	O
on	O
two	O
different	O
organs	O
and	O
diseases	O
;	O
breast	O
cancer	O
and	O
cerebrovascular	O
accidents	O
in	O
brain	O
,	O
commonly	O
referred	O
to	O
as	O
stroke	O
.	O
</s>
<s>
Similarly	O
,	O
the	O
MPRAD	O
features	O
in	O
brain	O
stroke	O
demonstrated	O
increased	O
performance	O
in	O
distinguishing	O
the	O
perfusion-diffusion	O
mismatch	O
compared	O
to	O
single	O
parameter	O
radiomics	B-Algorithm
and	O
there	O
were	O
no	O
differences	O
within	O
the	O
white	O
and	O
gray	O
matter	O
tissue	O
.	O
</s>
