Chapter 37: Neuroimaging in Psychiatric Disorders of Childhood

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement, not replace the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome to the Deep Dive.

Today, we're diving into research that really changed the game for understanding mental health in young people.

You know, for a long time,

even clinicians were pretty skeptical that kids could have serious psychiatric conditions.

It was a prevailing view, yes.

The idea that these were somehow adult problems or just behavioral issues.

Exactly.

But then came non -invasive brain imaging.

Over the past, what, 30 years, this tech basically opened up the black box.

It gave us the first real look inside the living, developing brain.

And suddenly, we could see clear biological roots for these disorders.

It wasn't just speculation anymore.

A total paradigm shift, like you said.

We moved from thinking these things only start later in life to actually mapping out how they developed, sometimes from very early on.

And that's our goal today, to give you a sense of the tools researchers use and some of the really quite surprising things they found.

Okay, so our mission, we'll explore the neuroimaging toolkit structure, chemistry function, then we'll get into the real world challenges of scanning kids' brains.

And finally, yeah, we'll unpack some key findings for conditions like ADHD,

OCD, and mood disorders.

Right.

So where do we start, I guess, with the leap from older methods?

Absolutely.

Before this, studying the biology was incredibly difficult.

Post -mortem brains, they tell you something, but they're confounded by age,

illness duration, treatments.

It's messy.

Not ideal for understanding development, for sure.

Not at all.

Now we can study the brain as it develops in vivo.

That's the revolution.

Okay, let's talk structure first.

MRI is the workhorse now, replacing things like CT scans because the detail is just so much better.

Much better resolution, much better contrast between different tissue types like gray and white matter.

Initially, though, people were just looking for obvious damage, right?

Like a stroke or a tumor, the radiologic approach.

Yes, but psychiatric disorders are usually more subtle.

There often isn't a big visible lesion.

So the field had to get smarter.

How so?

How do you measure subtle differences in, say, the gray matter?

We shifted to quantitative methods.

Looking at the cortex, the gray matter, which is where a lot of the processing happens, instead of just saying,

this area looks small, we started measuring specific things.

Like what?

Exactly.

Things like the actual volume of a region, its shape, the thickness of the cortex itself, the surface area, and even how wrinkly it is, that's called gyrification.

Gyrification, the folding pattern.

Exactly.

Cortical thickness, in particular, became a really key metric.

It often reflects how dense the neurons are, how organized that bit of cortex is.

Okay, so that's the gray matter, the processors.

What about the wiring, the white matter that connects everything?

How do we check the integrity of those connections?

For that, we rely heavily on diffusion tensor imaging, or DTI.

DTI, right.

Think of DTI as mapping the brain's highways.

It tracks how water molecules move.

In healthy white matter, you have these tightly packed bundles of axons running in parallel.

Like insulated cables.

Pretty much.

And water can't easily move sideways through those bundles.

It's forced to move along them.

We call that anisotropic diffusion.

It prefers one direction.

And there's a specific number that captures this.

Yes, it's called fractional anisotropy, or FA.

High FA means the pathway is well -organized, like a superhighway.

Water movement is very directional.

And low FA.

Low FA suggests the pathway might be disorganized, maybe damaged, or perhaps not fully myelinated yet.

The water can move more randomly in all directions.

So FA gives us a measure of white matter integrity, how robust those connections are.

That makes a lot of sense.

Okay, structure and connectivity.

But what about the brain's chemistry?

That seems crucial for understanding disorders.

It is.

But studying chemistry in kids brings up safety concerns pretty quickly.

You mean radiation, like with PETE scans?

Precisely.

Techniques like PETE and STECT are fantastic for looking at neurochemistry.

Like dopamine receptors.

But they use ionizing radiation.

That's a risk we try very hard to avoid in children, unless it's absolutely necessary for clinical reasons.

What's the alternative for research?

The go -to method is magnetic resonance spectroscopy, MRS.

It uses the same MRI scanner, just tuned differently so there's no ionizing radiation.

And what chemicals can MRS actually pick up?

Is it sensitive enough?

It's good for chemicals that are present in relatively high concentrations.

We typically use proton MRS, or 1HMRS.

We can measure things like N -acetyl aspartate, N -able A.

Think of that as a marker for how healthy and numerous your neurons are.

A neuronal health meter, okay.

Sort of.

We also measure choline, involved in cell membranes, creatine for energy metabolism, and importantly, key neurotransmitters like glutamate and glutamine, often combined as Glease and GABA, the main inhibitory one.

So excitatory and inhibitory balance.

Exactly.

But there's a catch.

MRS isn't sensitive enough to detect really low concentration chemicals, like the monoamines, dopamine, serotonin, or penephrine.

That's a limitation you have to remember.

Right, important limitation.

Now moving on to function.

This is where fMRI comes in, and it feels like this is where things get really dynamic.

Oh, absolutely.

fMRI was a huge leap because the resolution, both in space and time, is so much better than older functional methods.

How much better are we talking?

Spatially, we can get down to a millimeter or even less.

Temporally, we can capture brain activity changes happening roughly every second.

That kind of speed is vital for tracking thought processes.

And it works by tracking blood flow, right?

The BOL -D signal.

That's right.

BOL -D stands for blood oxygenation level dependent.

When a brain region becomes more active, it demands more oxygenated blood.

Deoxygenated hemoglobin has different magnetic properties than oxygenated hemoglobin, and the MRI scanner can detect that change.

So more signal means more activity, basically.

It's an indirect measure of neural activity, but yes, that's the principle.

Mapping those signal changes tells us which areas are working harder.

And it used to be you had to have the person do a specific task in the scanner, but that's changed too, hasn't it?

With resting state fMRI?

Yes, that's become incredibly powerful using advanced analysis methods like independent component analysis or graph theory.

Fancy math.

Yes, essentially.

But it allows us to look at the whole brain without a specific task.

We can see which brain regions tend to activate together forming intrinsic networks.

The resting state networks, even though the brain isn't really resting.

It reveals the brain's underlying functional architecture, how different areas are intrinsically connected and communicate even when you're just lying there.

It lets us explore connectivity patterns across the entire brain without needing a predefined hypothesis about one specific region.

Okay, so we have these amazing tools, structure, chemistry, function, but using them with kids?

That brings a whole different set of challenges.

Oh, enormous challenges.

Practical challenges, analytical challenges.

It's much harder than scanning adults.

Why is that?

What are the big hurdles?

Well, first, the analysis methods themselves are getting more complex.

Things like machine learning require huge amounts of data to be reliable.

We're talking studies needing well over 100 kids, ideally from multiple sites, to really trust the findings.

Getting that many participants is tough, but surely the biggest issue is just getting kids to stay still.

Motion artifact.

It's the bane of pediatric neuroimaging.

You've got this loud, tight tube, and you need the child who might be anxious or have ADHD, making it even harder to lie perfectly still for maybe up to an hour.

And you can't just sedate them.

Right, because sedation changes brain activity and chemistry, the very things you're trying to measure.

So you need clever workarounds.

Like what?

Well, the chapter mentions using a mock scanner.

It looks and sounds like the real thing, oh, without the magnets.

It helps kids get used to the environment,

desensitizes them.

Like practice.

Exactly, and there are lots of practical tips.

Table 37 .1 in the text lays some out things like detailed pre -skin training, using play therapy, letting them watch movies during certain scans, offering rewards or incentives.

It just takes a lot more planning and resources.

Beyond just keeping them still, there's another fundamental challenge.

The brain itself is changing so rapidly during childhood and adolescence.

The baseline isn't stable.

That's a critical point.

You're not scanning a static organ.

It's undergoing massive program changes well into the early 20s.

What are the main processes driving that change?

Two big ones, synaptic pruning and myelination.

Pruning is like weeding the garden, getting rid of weaker or redundant connections between neurons.

Making things more efficient.

Precisely.

And myelination is insulating the remaining important connections, the white matter pathways.

So signals travel faster and more reliably.

Both processes work together to sculpt a more specialized, efficient adult brain.

And we can actually see this reflected in the imaging data over time.

Oh, clearly.

Take gray matter volume.

It doesn't just increase steadily.

It actually peaks, hits its maximum volume somewhere in mid -childhood or early adolescence.

And then it goes down.

Yes, it then starts to decrease, reflecting that pruning process.

And interestingly, that peak happens earlier in females, around age 9 .5 on average, compared to males who peak around 10 .5.

Wow, a full year difference.

Roughly, yes.

And that reduction afterwards is thought to be really important for refining circuits and boosting processing efficiency.

Meanwhile, what's happening with the white matter, the connections?

White matter volume tends to increase more steadily, sort of linearly,

throughout childhood and adolescence.

This lines up with ongoing myelination.

And is that happening everywhere at once?

No, it's also regional.

The growth is particularly dramatic in areas like the dorsal lateral prefrontal cortex,

the DLPFC, which is heavily involved in executive functions like planning and decision -making.

This anatomical maturation underlies the cognitive maturation we see.

This idea of different regions maturing at different rates.

That's key, right?

It's called heterochronous development.

And understanding this normal timetable is essential before you can say anything about pathology.

And you mentioned sex differences in the timing of the gray matter peak.

That seems really important, especially since it lines up with puberty.

It does.

And it might help explain some sex differences we see in psychiatric disorders.

For example, look at the basal ganglia structures deep in the brain involved in movement and reward, like the caudate nucleus.

Okay.

Gray matter volume in the basal ganglia decreases more sharply in boys between ages six and 12 compared to girls.

And that timing.

That timing coincides with the peak age of onset, or diagnosis for disorders, heavily linked to basal ganglia dysfunction, like OCD and ADHD, which are more common in boys during that period.

Fascinating.

And what about girls?

Any similar links?

Well, think about the amygdala and hippocampus key structures for emotion processing and memory.

They also undergo significant changes around puberty.

Right.

And the timing of those changes seems to correlate with the rise in mood and anxiety disorders, which becomes significantly more prevalent in adolescent girls compared to boys.

So it's not just about having a disorder or not.

It's about how the brain's development diverges from the typical path.

Exactly.

Neuroimaging lets us pinpoint when and potentially how the trajectory for a condition like ADHD or depression starts to differ from typical development.

Okay, this sets the stage perfectly.

Let's dive into some of those specific disorders.

Starting with ADHD, attention deficit hyperactivity disorder.

The evidence here seems pretty strong.

It's one of the most studied areas.

And yes, the findings are quite consistent.

The prefrontal cortex, or PFC, which is crucial for things like attention, impulse control, planning.

It's consistently implicated.

How so?

Smaller?

Yes.

Structural studies repeatedly find smaller volumes in parts of the PFC, particularly areas like the superior and inferior frontal gyri in kids with ADHD.

But wasn't there something even more striking than just size?

Yes, the timing of its maturation.

Research suggests that the overall structural development of the prefrontal cortex in children with ADHD lags behind their typically developing peers by about two to three years.

Two to three years?

That's huge developmentally.

It is.

It provides a potential biological basis for why those executive functions, attention, impulse control, are developing more slowly.

The brain structure underpinning them is literally developing later.

What about other brain regions in ADHD?

The basal ganglia also show up frequently, often smaller volumes, sometimes differences in asymmetry compared to controls, and MRS, the chemical imaging.

Which is that?

Studies have found lower levels of NMLA, that neuronal health marker in the striatum, part of the basal ganglia, in ADHD,

and sometimes higher levels of GELX, which is mainly glutamate, the brain's main excitatory neurotransmitter.

So less healthy neurons and maybe too much excitatory activity.

That's one interpretation.

It could suggest reduced neuronal viability, possibly linked to an imbalance toward excitation in that key frontal striatal circuit.

And does treatment change any of this?

Can we see the effects of medication like methylphenidate?

Yes, there's evidence for that.

Stimulant medications like methylphenidate seem to help normalize some of those deficits in the frontal striatal pathways.

Functionally, FMRI shows that these medications consistently boost activation in areas like the right inferior frontal cortex and the insula.

Areas involved in cognitive control.

Exactly.

It suggests the medication helps the brain better recruit the circuits needed to manage attention and inhibit impulses.

Okay, let's shift gears to OCB, Obsessive -Compulsive Disorder.

Another disorder often starting in childhood.

What's the picture here?

The core idea in OCD revolves around dysfunction in a specific circuit.

The ventral prefrontal cortex connecting down to the striatum and then looping through the thalamus and back.

The corticostratothalamocortical loop, right?

That's the one.

And again, the chemical findings here using MRS have been particularly interesting, maybe even clinically predictive.

How so?

Studies found that children and adolescents with OCD who hadn't received treatment yet showed significantly higher concentrations of GEL -S glutamate glutamine, specifically in the caudate nucleus, part of that striatal loop.

Higher excitatory activity again, but you said predictive.

Yes, here's where it gets fascinating.

They follow these kids through treatment.

After 12 weeks on an SSRI antidepressant peroxetine, that high GEL -AX level in the caudate actually decreased back towards the level seen in healthy controls.

Okay, the drug normalized the chemistry?

It seemed to, but here's the key comparison.

In kids who received 12 weeks of cognitive behavioral therapy or CBT instead of the drug,

that normalization of BL -AX did not happen.

Wow, so the chemical change was specific to the medication.

In that study, yes.

And even more importantly, the kids who started with higher GEOX levels before treatment began actually showed a better clinical response to the peroxetine.

That's incredible.

It suggests that baseline brain chemistry could potentially guide treatment choices, maybe indicating who needs medication first.

It certainly raises that possibility.

It suggests that for some individuals, there might be a biological need to normalize that circuit chemistry pharmacologically, perhaps before or alongside behavioral therapy.

Are there other brain changes seen in OCD?

Yes, the thalamus, the relay station in that circuit also shows differences.

Some studies found larger thalamic volumes in pediatric OCD patients.

And again, after SSRI treatment, those volumes tended to decrease towards normal levels and that decrease correlated with how much their symptoms improved.

And this decrease didn't happen with CBT either.

Not in the same way, suggesting again, perhaps a drug -specific effect, maybe even a neurotrophic or structural remodeling effect from the SSRI.

They also saw the neuronal marker NAA increase in the thalamus after SSRI treatment, hinting at improved neuronal health.

It really paints a picture of medication potentially altering brain structure and chemistry in specific ways.

We should probably mention pandas briefly here too.

Right, the subtype of OCD thought to be triggered by strep infections.

It's a useful contrast.

Unlike typical early onset OCD where basal ganglia volumes might be normal or smaller,

kids with pandas often show enlarged basal ganglia.

Bigger.

Yes.

And critically, those enlarged volumes tend to shrink back towards normal size after treatment with immunotherapy, like IVE or plasmapheresis.

It highlights how imaging can help differentiate potentially different biological causes presenting with similar symptoms.

Okay, let's cover mood disorders, major depressive disorder, MDD, and bipolar disorder, BPI, in young people.

Where do we see differences here?

A key region that consistently comes up is the subgenual prefrontal cortex, the SGPFC.

It's deeply involved in regulating emotion.

And what's seen there?

Smaller volumes in SGPFC are often found, particularly in individuals who have a strong family history of depression or bipolar disorder, the high -risk cases.

What about the classic emotion centers, like the amygdala and hippocampus?

It's interesting.

Abnormalities there, like volume differences,

often seem to correlate more strongly with how anxious the child is rather than how depressed they are.

Anxiety is often comorbid, of course.

So amygdala size related more to anxiety symptoms?

Some studies found that yes.

For example, a larger ratio of amygdala volume compared to hippocampal volume was linked to greater anxiety severity.

Separately, smaller left hippocampal volumes have been linked specifically to familial, inherited forms of MDD.

Okay.

Let's briefly touch on a couple others mentioned, Tourette syndrome.

With Tourette's, you see findings pointing towards the motor pathways, things like subtle abnormalities in how the cortex developed, and often reduced volumes in parts of the basal ganglia involved in motor control, like the globus pallidus and putamen.

And dopamine seems key there.

Yes, PT studies, although involving radiation, have shown evidence of excessive activity of dopamine decarboxylase, an enzyme that makes dopamine in the caudate and midbrain, strongly suggesting a dysregulation to the dopamine system.

And finally, childhood onset schizophrenia, COS.

That sounds particularly severe.

It generally is much more severe than adult onset schizophrenia.

The brain changes seen in imaging are often more dramatic too.

We see a marked reduction in overall brain volume and enlarged ventricles, the fluid -filled spaces.

What's driving that volume loss?

It seems to be a progressive loss of both gray matter and white matter, particularly noticeable in the frontal and temporal lobes.

The idea is that it might represent an exaggeration of the normal synaptic pruning process, but going too far, leading to severely compromised brain circuits.

Wow.

Okay, we've covered a lot of ground from the techniques like MRI, DTI, MRS, FMRI, to the practical challenges, the importance of normal development as a baseline.

And then into specific findings across ADHD, OCD, mood disorders, Tourette's, and COS.

It really shows how these tools are providing biological windows into these conditions.

It's clear this field, while maybe still young in some ways, is already fundamentally changing how we think about these disorders.

Absolutely, and the future directions are pretty clear.

The big advantage of MR -based techniques is no ionizing radiation, so you can do repeated scans over time.

Longitudinal studies are crucial for tracking development and the effects of interventions.

Tracking kids over years.

Exactly.

And the other big push is for larger, more harmonized data sets across different research centers.

This is essential if we want to identify reliable biomarkers.

Biomarkers.

You mean things that could predict outcomes or guide treatment, like the Welex finding in OCD?

Precisely.

We need markers that are either moderators, present at the start predicting who responds best to what,

or mediators, things that change during treatment and explain how the treatment is working biologically.

That's the path towards more personalized medicine.

So let's leave our listeners with a final thought, going back to that ADHD, finding the two to three year lag in PFC development.

You said imaging can predict a child's age quite accurately just from brain structure, like cortical thickness.

Yes, the relationship between brain morphology and age is surprisingly strong in typical development.

So when you see a child with ADHD whose PFC looks structurally like that of a child two or three years younger,

they really are on a different developmental clock in that brain region.

It appears so.

Their brain development is literally off schedule compared to their peers.

Which leads to this provocative idea.

If we can map that schedule so precisely, could future imaging biomarkers do more than just diagnose?

Could they tell us when a particular child's brain might be most receptive, biologically speaking, to a specific therapy?

Could we potentially time interventions, maybe starting a medication or introducing a certain type of therapy to coincide with specific developmental windows identified through imaging?

Could we synchronize our treatments to the brain's own biological calendar?

Synchronizing therapy to neurobiology.

That's a really powerful idea for the future.

It's definitely where the field hopes to move.

Well, thank you for walking us through this complex landscape and thank you all for joining us on this deep dive into the neurobiology of childhood psychiatric disorders.