Information for Our Community

Whether you are part of our community or are interested in joining us, we welcome you to Washington University School of Medicine.

close  


Listen to more BioMed Radio episodes

Schizophrenia imaging

Researchers at Washington University School of Medicine in St. Louis recently reported they had discovered genetic evidence suggesting schizophrenia was not a single disease but a group of eight genetically distinct disorders, each with its own symptoms. Now, using advanced brain-imaging techniques, they’ve been able to match groups of patients to specific qualities in brain anatomy and function. The findings could represent another step toward improving diagnosis and treatment of the disorders commonly known as schizophrenia.

STUDYING CLUSTERS OF GENES, RESEARCHERS AT WASHINGTON UNIVERSITY SCHOOL OF MEDICINE IN ST. LOUIS RECENTLY FOUND THAT RATHER THAN A SINGLE ILLNESS, SCHIZOPHRENIA APPEARS TO BE SEVERAL DISTINCT DISORDERS. NOW, AFTER CONDUCTING BRAIN-IMAGING STUDIES, THEY ARE REPORTING THAT THE CONSTELLATIONS OF SYMPTOMS THAT CHARACTERIZE THE DISORDER ARE CLOSELY LINKED TO STRUCTURAL AND FUNCTIONAL DIFFERENCES IN THE BRAIN. JIM DRYDEN HAS THE STORY…

MANY PAST BRAIN-IMAGING STUDIES HAVE HAD A DIFFICULT TIME ISOLATING DIFFERENCES IN THE BRAINS OF PATIENTS WITH THE DISORDER, ACCORDING TO WASHINGTON UNIVERSITY PSYCHIATRIST ROBERT CLONINGER. HIS TEAM USED A TECHNIQUE CALLED THE GENERALIZED FACTORIZATION METHOD TO CLUSTER DIFFERENT BRAIN FEATURES TOGETHER.

(act) :21 o/c of schizophrenia

We were able to identify sub-groups that had been obscured when

all schizophrenics had been lumped together. These clusters

mapped right onto specific brain structures. The clustering

method actually retrieved well-known structures that have

functional significance in relation to understanding the

symptoms of schizophrenia.

FURTHER, CLONINGER SAYS, THE RESEARCH TEAM NOT ONLY FOUND SPECIFIC AREAS OF DIFFERENCE IN THE BRAINS OF PATIENTS WITH A SCHIZOPHRENIA DIAGNOSIS, THEY ALSO MATCHED SPECIFIC BRAIN FEATURES TO SPECIFIC TYPES OF SYMPTOMS, JUST AS THEY PREVIOUSLY HAD BEEN ABLE TO MATCH PARTICULAR CLUSTERS OF GENES TO SYMPTOMS,

(act) :16 o/c clinical syndromes

And so now we have a situation where we’ve shown that there are

distinct sub-groups among people with schizophrenia, and they differ

in terms of networks of genes, specific brain structures, and different

clinical syndromes.

RESEARCHER IGOR ZWIR SAYS TO IDENTIFY AND CATEGORIZE DIFFERENCES IN BRAIN STRUCTURES, THE RESEARCH TEAM USED A VERY SIMILAR METHOD WITH BRAIN IMAGES TO THE ONE THEY HAD PREVIOUSLY USED TO MATCH UP CLUSTERS OF GENES.

(act) :22 o/c previous paper

Instead of encoding results from genetic variants, we are now

encoding voxels. In the same thing that any image can be pixelized,

and you capture one pixel, this pixel here is now…or this voxel,

corresponds to one of the previous genetic variants that we had

before in the previous paper.

ONE KEY TO THIS METHOD, WHICH IS SIMILAR TO THE ALGORITHM USED BY NETFLIX TO PREDICT WHICH MOVIES A PERSON MIGHT WANT TO WATCH, IS THAT ZWIR SAYS THERE WAS NO ATTEMPT TO MATCH UP ANY BRAIN IMAGES WITH ANY SYMPTOMS. THEY WERE, HE SAYS, AGNOSTIC ABOUT THE BRAIN-IMAGING DATA.

(act) :13 o/c the data

We studied this information separately. So the “apples” and the

“oranges” are separate because otherwise, you are constraining

the information. One relationship is something that just comes

up from the data.

BUT EVEN THOUGH THE RESEARCHERS DIDN’T MAKE ANY ASSUMPTIONS TO TRY TO MATCH UP IMAGING DATA WITH TYPES OF SYMPTOMS, CLONINGER SAYS IT TURNED OUT THAT CERTAIN BRAIN DIFFERENCES WERE CONNECTED TO NEGATIVE SYMPTOMS, SUCH AS MONOTONE SPEECH OR AN INABILITY TO ACT SPONTANEOUSLY. MEANWHILE, OTHER FEATURES IN THE BRAIN WERE ASSOCIATED WITH SO-CALLED POSITIVE SYMPTOMS, SUCH AS DELUSIONS OR HALLUCINATIONS.

(act) :18 o/c as well

No forcing was done at all. He just analyzed the brain-imaging

data, and what emerged from that was one group with a lot of

positive symptoms and another group with very prominent negative

symptoms and others with bizarre behavior, and that was like

what we found with the genetics as well.

THE STUDY IS PUBLISHED IN THE JOURNAL NEUROIMAGE. I’M JIM DRYDEN…

RUNS 2:59