Imaging in Psychiatry

[FranzLO]

I just read a slightly older thread in the Rads forum and it was mentioned that a Psychiatrist can dabble in imaging if they choose to. I also wonder if EEGs and things can be read by a Psychiatrist and if specific training is required. I am pretty naive about this haha. This sounds fascinating to me, so I am kind of curious about this path. I have done some reading but I get a little confused about the different pathways you can take. I will most likely go the radiology route but I have an interest in mental health so this definitely sounds interesting. If I went the Psych route I wonder how this could fit in. Thanks!

---

Members don't see this ad.

 

[vanfanal]

I wouldn't expect this to be a part of your daily practice, at least not any time soon. I think most people would tell you that imaging in psych will mainly is still confined to research.

I agree with you in that I would be really cool and interesting to review imaging as a part of regular psych practice. Don't know much about EEGs to be honest.

 

[Frazier]

Agree with above. We had a psych luncheon a few weeks ago at at my school. One of the questions was [paraphrasing] "Where do you see imaging fitting into psychiatry"...

Both docs indicated that essentially it is an area of heavy research (at least at my institution -- fMRI, etc) and there is promise for some utility, but it is still years away.

Just to add on for completion, in regards to genomics and whatnot, they predicted that route will continue to be laid but are still decades away from definitive "tests" that might be utilized in the clinic.

If I was extremely visual-oriented and heavily interested in the prospect of reading imaging studies as a major facet of my career, I would not pursue psychiatry. You already know, radiology would be more likely.

 

[Shikima]

You can do a geri fellowship and integrate PET/MRI scans into your practice as a part of the evaluations for dementia related processes.

 

[shan564]

I think it'll be a reality during our careers. We already routinely get brain MRIs on people with first-episode psychosis, especially if it's an atypical presentation. They don't usually play heavily into the diagnostic/management process, but I've found them to be helpful on occasion. And there's a lot of research showing potential utilities in the future.

The chair of our department is an expert in electrophysiology, and he thinks that we're not too far from using EEGs either. I've never ordered an EEG unless I'm suspecting seizure activity, with the exception of ECT of course (although those EEGs are pretty routine/straightforward), but with noninvasive brain stimulation becoming a bigger thing, I'd be surprised if EEG didn't become a more routine monitoring test.

I think genetic testing is ready for clinical application already, but we don't utilize it because insurance doesn't pay for it yet. Anybody can get a fairly comprehensive genetic profile for a pretty reasonable price now, which was impossible even 5-10 years ago. Now that it's become exponentially cheaper and faster to sequence a genome, I don't see why we can't use the abundance of genetic data to apply to our patients for diagnosis, prognosis, and sometimes for selection of appropriate treatment (although pharmacogenomics is still in it infancy).

fMRI and PET would be useful if they were cheaper/easier, but I don't see insurance agreeing to pay for those things anytime in the near future, and nobody can afford to pay for it out of pocket. Personally, I like the idea of functional near infrared spectroscopy (fNIRS), which is far cheaper and easier but with some limitations.

 

[Merovinge]

It is possible to do a neurophysiology fellowship off of psychiatry (although quite rare), particularly one that can be tailored to be all EEG based and not EMG. Neuropsychiatry is a much more common fellowship and should leave you with some comfort level in reading brain MRIs. Also agree with above that over the course of our careers this will expand significantly.

 

[hamstergang]

We already routinely get brain MRIs on people with first-episode psychosis, especially if it's an atypical presentation.

I wish it were this way, somewhat. At my main hospital, we get CT scans for first break psychosis, despite the fact that MRI is the actually indicated test for what we're looking for, mostly. But even then, I thought the evidence suggested that MRI was not cost effective to be used routinely. Instead, I thought we should be obtaining a thorough history and performing a thorough physical/neurological exam and only get the MRI if there are abnormalities. Is this not the case?

The chair of our department is an expert in electrophysiology, and he thinks that we're not too far from using EEGs either. I've never ordered an EEG unless I'm suspecting seizure activity, with the exception of ECT of course (although those EEGs are pretty routine/straightforward), but with noninvasive brain stimulation becoming a bigger thing, I'd be surprised if EEG didn't become a more routine monitoring test.

Well, last year the FDA approved the use of EEG in assisting in diagnosing ADHD: http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm360811.htm

 

[SomeDoc]

To add to the EEG ADHD study;

"Carnegie Mellon University researchers have created brain-reading techniques to use neural representations of social thoughts to predict autism diagnoses with 97 percent accuracy."
http://www.cmu.edu/news/stories/archives/2014/december/december2_thoughtmarkersautism.html

Obvious disclaimer is that these emerging findings are good supplements, but not substitutes for a good clinical history, exam, and biopsychosocial considerations.

 

[shan564]

I wish it were this way, somewhat. At my main hospital, we get CT scans for first break psychosis, despite the fact that MRI is the actually indicated test for what we're looking for, mostly. But even then, I thought the evidence suggested that MRI was not cost effective to be used routinely. Instead, I thought we should be obtaining a thorough history and performing a thorough physical/neurological exam and only get the MRI if there are abnormalities. Is this not the case?

One of my colleagues diagnosed somebody with a CNS vasculitis based on an MRI. She just had little tiny lesions with no focal signs. I was convinced after that.

I still don't think that every single first-episode psychosis warrants an MRI if it's a typical presentation of schizophrenia or bipolar disorder. Just the slightly unusual ones. Sometimes the MRI will just show the classic changes typical of schizophrenia, which I also find helpful sometimes. But yeah, of course there's no substitute for a thorough H&P.

 

[st2205]

A better analogy is to look at neurology. They rely on imaging a lot. Sure, they'll look at stuff and can kinda read, but they're not radiologists, and if something is important that may alter management, they're consulting with the neuroradiologist and going with their final impression.

If neurology, being a field much more reliant on inaging, isn't reading independently (unless these guys do a fellowship), then I don't see us sitting down in the dark room (or on our desktop, really) looking at the inaging and saying, "Ah hah! There's our answer. We will augment with Remeron."

Now, I've had imaging be quite relevant, but it has nothing to do with psychiatry (i.e. "psychotic" [re: delirious] patient the ED wants us to admit that ends up having a subdural and massive midline shift). In those scenarios you won't need more than a radiologist -- or in the case above nothing more than a kid that can play "which of these is not like the other?"

 

[shan564]

When I rotated through neuro as an intern, it seemed like the neurologists were convinced that they could read a brain MRI better than a radiologist could. Or at least they would try to.

 

[Jorje286]

IMO it's still a long way to go. It's one thing to establish 'biomarkers' for diseases diagnosed with DSM, and it's another thing to diagnose and classify diseases based on such biomarkers.

 

[smalltownpsych]

We probably need more effective medications that can target specific areas of the brain as opposed to the shotgun effect of medications that we use now. Otherwise having more detailed diagnostic information from imaging is not very useful. I really liked that info about the part of the brain responding to the processing of social information. We do need to know more about that and how that is tied into learning and affect regulation.

 

[splik]

Sometimes the MRI will just show the classic changes typical of schizophrenia, which I also find helpful sometimes.

do you know something the rest of us don't?

 

[shan564]

do you know something the rest of us don't?

I'm just talking about the nonspecific stuff like ventriculomegaly, which doesn't have a very wide differential when combined with a 22-year-old patient presenting with vague psychosis.

 

[notdeadyet]

You are also not likely to find it in a 22yo with early vague psychotic symptoms. An MRI for diagnosis of psychosis has such low sensitivity.

Brain imaging in cases of psychosis can be helpful to rule out neurological causes of the disease. We don't order them as a rule in for schizophrenia.


Sent from my iPhone using Tapatalk

 

[shan564]

You are also not likely to find it in a 22yo with early vague psychotic symptoms. An MRI for diagnosis of psychosis has such low sensitivity.

Brain imaging in cases of psychosis can be helpful to rule out neurological causes of the disease. We don't order them as a rule in for schizophrenia.

I agree, but I used that example because it was real on two occasions in my intern year. The patients were both 21ish-year-old females with somewhat vague presentations. One of them looked bipolar, the other one looked like she might be severe MDD and/or personality. Both had big ventricles and small temporal lobes, which helped us decide on the diagnosis of schizophrenia/schizoaffective. Both were re-diagnosed with something else (incorrectly, in my opinion) on a subsequent admission (substance-induced psychosis and MDD, respectively), partially because they didn't look at the MRI. The MDD diagnosis was made just because her psychotic symptoms were under control with Sustenna, so they didn't see any psychosis. The substance-induced diagnosis was because she decompensated after smoking a lot of weed, but I saw her psychotic in the absence of substances.

Either way, the diagnosis could have been made correctly without brain imaging, but I thought it helped us decide sooner.

I still don't think that brain MRI is good enough to rule in or rule out schizophrenia, but it might be helpful in some cases. It was much more meaningful for that girl who turned out to have vasculitis.

I don't think there's enough evidence to say that it's "right" or "wrong" to get a brain MRI, even for an atypical presentation.

 

[hamstergang]

I agree, but I used that example because it was real on two occasions in my intern year. The patients were both 21ish-year-old females with somewhat vague presentations. One of them looked bipolar, the other one looked like she might be severe MDD and/or personality. Both had big ventricles and small temporal lobes, which helped us decide on the diagnosis of schizophrenia/schizoaffective.

Do you know the sensitivity or specificity for ventriculomegaly on MRI indicating schizophrenia?

 

[MacDonaldTriad]

The 90s where declared the “decade of the brain” by the NIH and this coincided with the advent of functional imaging. I started my research career in functional imaging. I bought my first house and buckled down to work. Now my mortgage is almost paid off and we still do not have much in the way of clinically relevant imaging in psychiatry. I’m sure there will be a lot of people who will give examples of cases where imaging was the key to what was happening, but it still isn’t routine.

 

[notdeadyet]

I don't think there's enough evidence to say that it's "right" or "wrong" to get a brain MRI, even for an atypical presentation.

Agreed. Not right and wrong. But there is "standard of care."

Even in psychiatry. Much like how often and at what age women should get mammograms or when you should give antibiotics for head colds. You can deviate and it's hard to say that it's "wrong." And folks will inevitably have annecdotes about how it was "right." It's just not standard of care.

 

[hamstergang]

One of them looked bipolar, the other one looked like she might be severe MDD and/or personality. Both had big ventricles and small temporal lobes, which helped us decide on the diagnosis of schizophrenia/schizoaffective.

Also, here's a study showing ventriculomegaly in bipolar disorder: http://psychiatryonline.org/doi/abs/10.1176/appi.ajp.159.11.1841

So how can this finding help differentiate between bipolar disorder and schizophrenia? I feel like MRI in this case can be misleading, and therefore I think it is ok to call it wrong.

 

[vistaril]

I think it'll be a reality during our careers.

And who is going to teach all these psychiatrists how to read imaging?

Psychiatry has its plusses and minuses......but the minuses have the potential to outweigh the plusses when people keep trying to make it something it's not.

 

[NickNaylor]

The 90s where declared the “decade of the brain” by the NIH and this coincided with the advent of functional imaging. I started my research career in functional imaging. I bought my first house and buckled down to work. Now my mortgage is almost paid off and we still do not have much in the way of clinically relevant imaging in psychiatry. I’m sure there will be a lot of people who will give examples of cases where imaging was the key to what was happening, but it still isn’t routine.

I wonder if this is a return to the first histologists that were completely convinced that the psychiatric disease was a biological phenomenon that could be understood by looking under the microscope given enough exploration. That ended up not being the case, but with the advent of more advanced imaging techniques there seems to be some hope that, aha!, this will be the holy grail. Maybe it will, maybe it won't. I dunno. I know that there are correlations between abnormal structural findings and psychiatric processes, but this doesn't seem to be fully accepted.

I'm not so convinced that imaging will be the holy grail some are expecting, but that doesn't mean that it will be necessarily useless in a diagnostic context. It just seems too early to tell with relatively too little research to support that kind of approach.

 

[Jorje286]

I wonder if this is a return to the first histologists that were completely convinced that the psychiatric disease was a biological phenomenon that could be understood by looking under the microscope given enough exploration. That ended up not being the case, but with the advent of more advanced imaging techniques there seems to be some hope that, aha!, this will be the holy grail. Maybe it will, maybe it won't. I dunno. I know that there are correlations between abnormal structural findings and psychiatric processes, but this doesn't seem to be fully accepted.

I'm not so convinced that imaging will be the holy grail some are expecting, but that doesn't mean that it will be necessarily useless in a diagnostic context. It just seems too early to tell with relatively too little research to support that kind of approach.

It has to be "biological". If not, then that makes one a dualist, which imo is not a philosophically or scientifically tenable position. One of the biggest misunderstandings of the whole debate is that arguing that the brain is the seat of behavior somehow means that things like social factors, empathy or even therapy are useless and it|s all about neurotransmitters and genes. That's not really what those on the brain team say. Ultimately interpersonal and social factors will leave their imprints on the way brain functions.

The relevance of imaging though is a different question. I'm not so optimistic either that this will revolutionize the field. The way things are is basically like using a loupe to understand the way a cell functions. Every voxel contains a huge number of cells, and they are probably doing very different things, which may also not be simply accounted by how much blood gets to them. Our tools to understand the brain are still way too primitive. If I want to put my 2 cents on something it will be on computational and theoretical understanding of the brain. We need very complex theory to start putting a framework of how things could work in the brain before we actually try to experiment. And with psychiatry we are really talking about the most complex, most intircate level of brain functioning.