Biomedical Brain Imaging Advances in Clinical Research: Novel Uses of FMRI, MSRI, and PET.mp4

April 11, 2022

ID7687

To CiteDCA Citation Guide

00:00We're now going to hear three
00:04exciting short presentations that
00:06are from the what was one of our
00:09main research modules that highlight
00:13Yale strengths and neuroimaging,
00:16which is for which Yale
00:18is known internationally.
00:19So our first presenter
00:21is Doctor Todd Constable,
00:23who is Vice chair of Radiology
00:26and and heads component of the
00:29MRI Center. So Todd, please.
00:37Thank you John.
00:40I don't have it explicitly in my slides,
00:42but thank you to the Reverend and Tisha
00:45for that insight about the master program.
00:47I will touch on a few places where we've
00:52benefited by having a more diverse study
00:56sample and it's it's super important
00:58we going to talk about a paper briefly
01:01that we have under review at Nature
01:03where we point out that the brain shows.
01:08We have evidence from our analysis
01:12that when cognitive tasks were not
01:15designed for diverse populations,
01:18we can actually see that
01:19in the brain imaging data,
01:20so we can reveal some of these flaws
01:23and imaging studies as a result of
01:27not having the correct cognitive,
01:30culturally sensitive measures.
01:32So I'm representing both the
01:34pet center and the MRI.
01:36Center within the MRI center.
01:37I direct the MRI program.
01:39Doug Rothman directs the Mrs
01:40program and Rich Carson directs
01:42the pet program the MRI centers.
01:44Here in this annelyn center,
01:46and we've been in there since 2003.
01:49What am I?
01:53There we go.
01:53So within the MRI center I'll
01:55I'll talk about each group,
01:56kind of separately within the MRI center.
01:58We have 3 Siemens Prisma scanners.
02:00They do a lot mostly F MRI but
02:02they do a lot of other stuff too.
02:04We just recently acquired a 3T wide,
02:07Boris Siemens and this will really
02:09help people in an internal medicine
02:11who are doing obesity studies.
02:13And we also have a 4T Brooker system
02:15for Mr spectroscopy studies we have
02:17three what we call animal magnets for
02:20preclinical work at 14 nine point 411.7.
02:23Tesla and these higher field
02:24strains that allow us to go to
02:26higher and higher resolution,
02:28particularly important when imaging
02:29a really tiny brain like a mouse,
02:30brain, and so in.
02:33These preclinical models we've
02:35also developed in conjunction
02:36with folks in neuroscience,
02:39optical imaging,
02:39so we can do simultaneous kind
02:42of mesoscopic optical imaging
02:44and MRI within the MRI center.
02:47We have all the computing infrastructure,
02:49data transfer and support mechanisms needed,
02:52and so we think we have a.
02:54A system and a set up such that
02:56people that want to incorporate
02:58imaging into their research should
03:00be able to thrive specifically now,
03:03under Doug Rothman's direction there,
03:06there's a very active Mr Spectroscopy
03:08program where there are 10 faculty and
03:11research scientists who develop new methods,
03:14and I'm going to highlight a couple of
03:17recent developments that are important
03:19and other people are starting to adapt in.
03:22In all of these centers in the MRI,
03:24the Mrs.
03:24And the pet center.
03:26We develop new tools and then
03:28we we have outreach to look for
03:31clinicians and scientists to apply
03:33those tools to clinical populations,
03:36which is our ultimate goal.
03:37So in the Mrs Group.
03:40They mostly do their work at 4T
03:42on the the human broker system.
03:44There's also some preclinical work done,
03:46but right now we're in the process
03:48of translating some of those things
03:50to the more common 3T systems,
03:52which we actually have both
03:53clinically and in the MRI center,
03:55and so those that will make some of
03:57these spectroscopic methods more accessible.
04:00The Doug's lab has a metabolic
04:02modeling core and a biochemistry
04:04core for tissue sample preparation.
04:07Here's 3 examples of kind of
04:09innovative work that's been.
04:10Under under way for some time now
04:13in Proton spectroscopy you can
04:16do a motor cortex spectroscopy.
04:18Let's say from this region here or
04:22posterior cingulate, and you can get
04:25the proton Spectra and from editing
04:27that you can get quantifiable measures
04:30of lactate, glutamate and GABA.
04:34And by doing this you can really look at the
04:37metabolic consequences of activation and or.
04:40Just you know, even brain, resting state,
04:42and so this provides sort of quantitative
04:45insight into the local biochemistry
04:46that's going on in the brain.
04:48You can do this dynamically, so there's C.
04:5113 and fusion studies.
04:53These are almost analogous
04:54to FDG pet in some ways,
04:57and that you can look at
04:58the uptake of glucose.
04:59You can measure glutamate
05:01and glutamine quantitatively,
05:03and here's a study where they're looking
05:05at the effects of ketamine on the brain
05:08and how that changes compared to a placebo,
05:10and also.
05:10How that changes as a function of
05:13dose here and finally something
05:14that's having a large impact is
05:17again looking at the sorry this is.
05:19This is also analogous to FDG PEN,
05:23but you can do this with deuterium
05:26NMR and this is the looking at the
05:29Warburg effect in cancer and that's
05:31the process where there's active.
05:35Glucose metabolism in in the tumor region,
05:38but it produces excess lactate and
05:40they can directly get an image of this
05:42lactate and you can see it provides
05:43slightly different contrast to what you
05:45would get in a conventional anatomic Mr.
05:48Image,
05:49and Zachary Corbin is going to
05:51discuss this in detail in the
05:52third talk in this session.
05:56Moving on to the MRI resources,
05:58that's the part that I direct.
05:59We have programs in cardiac Mr real
06:02time feedback, cancer imaging,
06:04preclinical imaging and hardware
06:06and engineering development and
06:08just onto a back to the reverends
06:11and and tisha's comments in
06:12terms of Community development,
06:14one of the hard I'm not going
06:15to talk about it here,
06:16but one of the hardware engineering
06:18things that we're working on is
06:20developing low cost MRI scans that
06:22can be MRI scanners that can be
06:24put in the Community and used.
06:26Much more made, much more accessible,
06:29and that's kind of analogous
06:31to how ultrasound is used.
06:3480% of the world right now doesn't
06:36actually have access to MRI,
06:37and so we hope to be able to
06:39change that with some of the
06:41developments we're working on.
06:43One of the main things
06:44that we've got a lot of.
06:47Headway on is developing methods
06:50to link brain organization.
06:52This is primarily talking about
06:54functional organization to behavior and
06:56we can develop models that link brain
06:58to behavior and using these models
07:00we end up identifying the systems,
07:02supporting behavior and so this
07:04has important implications in say,
07:06psychiatric or neurologic studies
07:08where you want to see what the
07:11what the systems are that are
07:14contributing to symptoms, let's say.
07:17And so in the Ardoch formalism of NIH,
07:20there's a cognitive.
07:21There are cognitive constructs
07:22that are well defined,
07:23and you can model cognitive constructs,
07:26and you can also model symptom scores,
07:28and by doing this we can collect fMRI data.
07:31We build a map of all the
07:33connections in the brain,
07:34and these connections for or
07:36something like a 268 note Atlas.
07:38These there's 35,000 connections.
07:40There's a lot of information in
07:42these connections about the subject,
07:44the individual or the patient,
07:46and we're just learning how to read.
07:47That now so we can stack these
07:49across groups of individuals and the
07:51more diverse this population is,
07:53the better these models are that
07:55we can build,
07:56and we can then correlate or
07:59relate some sort of behavioral
08:01trait or symptom scored for each
08:05individual to identify the systems
08:07which is identified here that vary
08:10as a function of performance on
08:13a task or some symptom score,
08:15and these are actually predictive models,
08:17so these are not.
08:19The associations were able to predict
08:21left out individuals or independent
08:23groups and predict their behavioral
08:25scores from their imaging data,
08:27and so this is we can look
08:29at a range of traits we have.
08:31We're we're establishing a library right now.
08:3416 measures, cognitive constructs,
08:36and then probably another 16 symptom scores,
08:40and we can, you know,
08:41we get those for each individual.
08:43We can build these models and these
08:45models then tell us the systems that
08:48are responsible for supporting that.
08:50Behavior or that spectrum of behaviors,
08:52and we want to develop normative
08:54Spectra for that,
08:55and then see where patients individual
08:57patients lay on that spectrum.
08:59So the networks defined here
09:01reveal the systems and assessing
09:03then who the models fail,
09:04for whom the models fail is actually
09:06a way to kind of subtype people.
09:09So who has different functional
09:11organization brain behavior relationships
09:13such that the model doesn't fit them and
09:17we've had tremendous success with this?
09:19A bunch of nature.
09:20Papers in the last five or seven years we've
09:24got a lot of diverse labs involved here,
09:27so Michael Prayers Lab just
09:29Carden and Mike Higley,
09:31Marvin Chun over in psychology we've had,
09:35we've made great progress with this,
09:37and we actually have a nature
09:38paper pending right now,
09:39which is on the subtyping and for
09:41whom the model fails sort of analysis.
09:44Over in the pet Center,
09:46Pet center is very large.
09:47There's three cameras,
09:48pet cameras over there.
09:50This is a cyclotron,
09:51so with the cyclotron does is,
09:53it accelerates atoms?
09:54Very high energy and then smashes them
09:58into a target and creates radioisotopes.
10:01And this is a chemistry module that
10:04then puts those radioisotopes onto a
10:07ligand that can be injected in an individual.
10:09And then you can see where it goes
10:11and you get images like this.
10:13This is the uptake of radio.
10:14Racers and this is like mapping
10:16glutamate receptors in the brain,
10:18for example,
10:19and the the pet center develops
10:22these these ligands and each
10:25they're constantly developing new
10:27and novel targets that researchers
10:30within the university can then use
10:33so they have 12 hot cells.
10:35They have three of these whole
10:37whole well one head system and
10:39two whole body pet scanners,
10:41and then they have 3 little micro.
10:44Pets,
10:44pet scanners and a pet CT scanner for
10:47preclinical studies and you'll see
10:49in a minute how those come into play.
10:52So right now there are 62
10:54different radio tracers available.
10:56There's 160 radio tracers used
10:57in animals and a number of these
11:00ones that are being developed in
11:02animals ultimately do get translated
11:04to use in humans.
11:05There's 70 NIH grants using the
11:07Pet center and 50 human protocols.
11:10They're ongoing 3 areas of developmental
11:12briefly touch on before I wrap up.
11:15Is a synaptic density imaging this
11:18is a way that they can target the.
11:22The synapses in the brain and get maps.
11:25It's almost like a Gray matter map,
11:27reflecting, reflecting synaptic density,
11:29and they've already so that Yale
11:31wasn't necessarily the first to
11:33develop this for their first to
11:35have a really good ligand for
11:37this and make it practical.
11:39And so they've been able to make
11:41tremendous headway in in developing
11:43this and looking at specific diseases.
11:46So in epilepsy, Alzheimer's,
11:49and all sorts of psychiatric.
11:52Illness so there's over 20 grants
11:55on this already, and you know,
11:5740 L publications.
11:58But Yale is kind of a leader
12:00in this and the next talk,
12:02actually by Doctor.
12:03You know Esther Liz is going to
12:05actually link some of those SP2
12:07imaging to the connectivity.
12:08Imaging that I was talking
12:09about earlier with fMRI.
12:10So we're doing multimodal
12:12studies as well in the pipeline.
12:15Right now there's some novel tracers that
12:17are coming out for human work,
12:19so these are brand new.
12:20They they haven't really been used yet.
12:23Kappa Kappa receptors musical
12:27security golden genic receptors
12:30and there's also some tracers.
12:35I missed one.
12:38There's also some tracers in preclinical
12:41studies right now being developed in in
12:44animal models and those will be hopefully
12:46be ready for human use in a few years.
12:50Finally, there's a new pet
12:51scanner coming which is going to
12:53have much higher resolution,
12:55and it's going to have spatial resolution
12:57of the order of two millimeters voxel size,
12:59and that's comparable to what we get in F,
13:02MRI, and so when we're linking
13:04functional MRI and PET studies together,
13:07we'll have comparable resolution,
13:09which is great, and so this is being
13:12developed in collaboration with
13:13Rich Carson's group here at Yale,
13:15UC Davis in United Imaging,
13:17and that'll again give us kind of the latest,
13:19greatest technology.
13:21For doing research,
13:23these are the faculty and the MRI center.
13:25That kind of developed these methods and
13:27and support a lot of the infrastructure.
13:29And then I didn't download
13:31just the faculty for PET.
13:33But here's all the people and
13:34the Pet Center faculty and staff,
13:37and so there's a large cohort
13:39of people that come together to,
13:42you know,
13:42develop these new methods,
13:44and then we're always looking for partners
13:46and outreach and applying these clinically.
13:49So thank you.