MGH Martinos Center for Biomedical Imaging 2017 | Page 36

can target Type I collagen fibers.
Caravan is an associate professor
of radiology at Harvard Medical
School, a senior faculty member
in the Center and senior author of
the Hepatology paper. How can
this help in imaging liver fibrosis?
Collagen fibers make up much of the
excess connective tissue deposited
by fibrosis, so imaging the collagen
itself means researchers and clini-
cians can more directly probe the
molecular mechanisms of fibrosis.
This in turns allows them to detect
fibrosis in its earlier stages.
capabilities—collagen imaging was
most useful in distinguishing early
fibrosis; MRE, in distinguishing
advanced fibrosis—to come up
with a better means of staging liver
disease. “Our findings allowed us to
develop a composite fibrosis staging
metric, utilizing data from both
techniques, obtained in a single
imaging exam, which demonstrated
superior discrimination across all
stages of fibrosis progression,” Zhu
says.
Because of its potential for advanc-
ing the detection and staging of
The researchers initially set out to liver fibrosis—an important clinical
simply compare and contrast the need—the researchers are explor-
effectiveness of the two methods ing the possibility of developing
but soon realized they could take the combined approach for use in
advantage of their complementary clinical settings. MRE has already
received FDA approval so the
primary hurdle to be overcome
is approval of the gadolinium-
based agent. A company, Collagen
Medical, is now pursuing this com-
mercially.
At the same time, the Martinos team
is planning further studies with the
approach, exploring the ways it can
help in other models of liver disease.
“We are now looking in a model
of non-alcoholic steatohepatitis,
or NASH, which is a associated
with the epidemics in obesity and
diabetes,” Caravan says. “NASH is
a cause for great concern these days
because it can lead to liver failure,
cirrhosis and primary cancer of the
liver.”
New Software Improves Ability to Determine
the Cause of Stroke
Determining the cause of an
ischemic stroke is critical to prevent-
ing a second one and is a primary
focus in the evaluation of stroke
patients. But for all the importance
of identifying the cause, physicians
have long lacked a robust and objec-
tive means to do so.
sifications systems are often used
in research and clinical practice,
these systems are not always able
to produce subtypes with discrete
pathophysiological, diagnostic and
prognostic characteristics,” says
Hakan Ay, a vascular neurologist,
Martinos Center investigator and
senior author of the JAMA Neu-
Now a team of investigators at the
rology paper. “We found that the
MGH Martinos Center for Biomed-
CCS-based classifications provided
Hakan Ay
ical Imaging and the MGH Stroke
better correlations between clinical
Service have developed a software
and imaging stroke features and
package that provides evidence- (CCS)—was published online in were better able to discriminate
based, automated support for diag- JAMA Neurology.
among stroke outcomes than were
nosing the cause of stroke. Their
two conventional, non-automated
study validating the package—called “This was a much-needed study classification methods.”
Causative Classification of Stroke because, although stroke clas-
33