Wired Differently: Understanding Neural Networks in the Brain

Video Transcription

Steve Kerschke: Yeah. All right. It’s 9:00 AM. Those of you who are still filtering in welcome and those of you who’ve been sitting around listening to Yap for five minutes, you know that we have a couple special guests in studio today. Dr. Jeff Snell, a regular on the podcast/webinar. Welcome back. Always glad to have you. And then if you’ve been around a while, you also recognize the name Brad Dexter. So Brad, we’re very glad to have you back. We stole you back from UNMC even if it’s just for an hour. But Brad-

Brad Dexter: Happy to be here.

Steve Kerschke: Yeah, we’re glad to have you. So glad to

Brad Dexter: Have you here.

Steve Kerschke: Yeah,

Brad Dexter: Absolutely.

Steve Kerschke: Brad, how long have you been there now?

Brad Dexter: Two years.

Steve Kerschke: Two years. Two years ago, parted ways with QLI amicably and is fulfilling I think one of your dreams to be a teacher at UNMC. I’m very jealous of your students because I’m sure one of your classes are very well attended. At any rate, we’re going to be talking about neural networks today. This is a follow-up presentation from our, I think we did FME.

Jeff Snell: Yeah, this came out of I think our vestibular presentation.

Brad Dexter: Vestibular. And then we did functional neurological disorder last fall. I think it was last fall.

Steve Kerschke: And this was a follow-up to that and I think very just closely related to a lot of the concepts. So if you were around for that, some of this will be a bit familiar. If you weren’t, reminder, all of our presentations are hosted on YouTube. That one might be one to go revisit if you happen to miss that. So Jeff, we’ll take care of some housekeeping, everyone’s favorite part.

Jeff Snell: Okay.

QLI Overview & Housekeeping

Steve Kerschke: Just a reminder, we’re representing QLI today. QLI is a post-acute rehab facility. We also have an ever growing outpatient telerehab program, but at any rate, we are always focused on TBI, SCI. We have a small chronic pain program, limb loss and stroke. So we specialize in anything neurological or complex. You can see our campus there is on the right hand side. We sit on about 60 acres, half of it is developed. We have some really cool projects going on. One of them being our collective, which is more of a community facing project. If you’re interested in that, we’d love to tell you about that. If you are a little less familiar with QLI and would love a virtual tour, let Don or I know will take care of that too. And man, if you’re in Omaha, College World Series is coming up. If you want to stop by QLI, we’d love to host you.

So shameless plug. This is our mission. It’s something that we talk about a lot and at really every level … You want to hit Roseanne’s question. Really at every level of the organization, our mission is something that we discuss, but also hopefully something that people experience and it’s a big part of who we are and what we do. And then we all know that your CEUs are very important to you. We will get you those. As always, we just ask that you spend the majority of the time with us today. And upon clicking leave webinar, you should get the opportunity to submit feedback. That feedback is important to us mainly so that we can give our presenters positive and any sort of constructive feedback, but then it also allows us to provide you the free CCMC CEU. So make sure you do that. If you have issues, sometimes it’s your popup blocker.

Typically, it’s on your end, not ours. At any rate, Don and I will help you out if needed, so just reach out on that. And then throughout the presentation, please use the Q&A for any questions. We will monitor that along the way. Brad and Jeff will either tackle it as it comes up or we’ll hit them at the very end. With that, I’ll turn it over to you two.

Brain Regions Vs. Neural Networks

Jeff Snell: Excellent. Okay, Brad. Let’s go. Yeah. All right. What we’re going to be talking about today, brain regions versus neural networks. I mean, neural networks is the topic.

Steve Kerschke: Slide it over. Just make sure it’s in front of you.

Jeff Snell: Oh, it’s going to block my face. Nobody’s going to be able to see it. Okay. Slide it down.

Brad Dexter: Did someone just do applause?

Jeff Snell: I think so. Yeah. We got a face for radio, right? You heard that one? Okay. We’re going to talk about in kind of a superficial manner, several different neural networks. I don’t want to get too far into the weeds. I’m bad about doing that at times, but I’m going to try to keep this at least enjoyable as we talk about some of these networks and all the different parts and pieces involved. And why are these particular networks as with all aspects of the brain vulnerable to injury, particularly from the standpoint of traumatic brain injury aspects of that, but also other aspects of different processes by which our brains can be compromised. And then Brad is going to go a little bit more in depth into some of the more challenging clinical cases that are manifested as a result of disruption within these neural networks, as well as recommendations and resources for what is an appropriate empirically based treatment for such disorders.

Brad Dexter: Yeah, that’s great. A couple things I would add. I think I was thinking back to the FND webinar that we did last fall and one of the concepts that we talked about at length in there was hardware versus software. And you guys will probably get a sense of that as Jeff starts diving into brain regions as opposed to neural networks. The concept of hardware was there’s something, some part of a system that just needs to be replaced or fixed, right? It needs to heal lab whereas software, it’s like it needs an update of some kind. And so Jeff, you can run with that analogy as much as you want to, but I’ll turn it back to you now.

Jeff Snell: Well, I think the other is the brain is often referred to as wetware

Because computers don’t have the capability of fixing themselves when something goes wrong with them. Our brain actually has the ability to not just reprogram the software but to rebuild the hardware and that neuroplasticity is a part of that process. And so that’s why the brain is often called wetware as opposed to hardware and software. And because the brain is capable of changing itself in response to experiences and other physiological events as well, our brains have the ability to change positively or change negatively. And so some of the things you’re going to be talking about are reflections of those negative neuroplasticity adaptations that can occur. So we’re going to be talking about regions versus neural networks. I think most of the time what we’re talking about when I teach the brain historically has involved regions rather than networks. So what we’re most familiar with on a grand scale are the cerebrum as you see kind of pictured here are the cerebellum as kind of broad regions.

We often talk about the distinct lobes of the brain, frontal, parietal, temporal, occipital, and their responsibilities and functions. And within these lobes, there are also sub-regions that are identified by anatomical distinctions or associated functional processes. So this is kind of a classical model of assigning a singular anatomical area being responsible for a given function. So at the very back of the frontal lobe though the motor strip with the primary motor cortex, that pre-central gyrus, you’ve got a sulcus that goes across kind of ear to ear on your brain, which represents the back of the frontal lobe and the beginning of the front of the parietal lobe. Those are the motor cortex and the sensory cortex. The motor strip is responsible for sending out motor signals to the muscles in our body regulating that so that we are able to move about to consciously engage in action.

But in reality, there are a number of different steps along the way. So it’s not just the motor strips. Kind of like when you talk about the frontal executive function residing within the frontal lobes, if other parts of your brain are not functioning very well, but your frontal lobe’s perfectly intact, it doesn’t necessarily mean you’ve got great reasoning and judgment, but these other things are compromised because all of that is interconnected. We talk about the fact that all these different parts and pieces of the brain have to communicate with each other in order to achieve the functions. And so as a result, even when we’re talking about regional parts of the brain as opposed to neural networks, these are still interconnected. They’re interdependent and as a result, damage in any area can also affect these. We also talk about regions in terms of physiological structures within the brain.

Micro, Meso, And Macro Networks

For any of us that have ever been through any anatomy course, I wish, I wish that these structures were so conveniently color coded like they are in this image. It all looks the same once you get inside there, but there are structures within the brain that are identified that have certain functions. The corpus callosum that communicates from one side of the brain to the other, the hippocampus that’s so involved in memory functioning and emotional regulation as well. These all have to interface with everything else. So all of these different parts and pieces when we’re talking about regional aspects are things that represent how the brain functions and where major functions come from within the brain. Now, neural networks, on the other hand, relate and reflect different structures within the brain or clusters of brain cells that interact with each other and communicate with each other in order to achieve a specific function, a specific outcome or an action.

Now, this can be on a micro level where small localized groups of neurons perform a very specific and specialized task within the brain. Examples of micro levels of circuits within the central nervous system would be reflected. For example, in this diagram, which is Rinshaw cell circuits in the spinal column, they provide an immediate little feedback loop to a spinal motor neuron or within the cerebellum where a Purkinje cell actually interfaces with perhaps even a hundred thousan synapses from the parallel fibers, but it is a single cell microsystem where information is being accumulated to determine if an inhibitory signal needs to be sent. Another is just within the cortex of the brain. Here are five columns represented of the six different cell layers that make up the majority of the outer cortex of our cerebrum. Each of these columns would be associated depending on where they are located within the brain on a certain type of function.

So if these are in the frontal lobe, depending on where they are, they might have an inhibitory function or excitatory function. They might be a gating mechanism. So within different parts of the brain, these layers of columns of cells are producing a signal and it’s a very micro level of functioning. As you build these circuits together, you go to then the next level, which is meso networks. These are in between micro and macro levels, meso means in between. These all come from Greek word roots to begin with. Meso networks would be reflected by something like the hippocampal circuitry of information being passed between three major cell groups within the hippocampus. So this is a mesocircuit that forms crucial components that are effective when it comes to memory consolidation and encoding. It also is very simple to spatial navigation interestingly enough as well. So these even at the meso level can be very differentiated into different functional abilities within a single regional area.

Also, for example, our olfactory system. So this is a group of cells that communicate with each other and basically form a gating mechanism between the receptors versus what actually then gets sent to our conscious awareness of our sense of smell. So these are in between the micro level and the macro level. Expanding further out though, generally encompassing large scale whole brain networks. And that’s what we’re going to be talking about today are the macro networks. So these networks coordinate complex processes, complex behaviors, cognitive communication processes, motor and sensory actions, our perceptions. These macro networks incorporate information from the underlying micro and meso networks and then coordinate that throughout the brain for necessary information processing, decision making, goal making behavior, goal directed behavior. All of the physical, cognitive and emotional processes that our human brain is capable of are represented within these different neural networks. Now, what you see picture on the screen here involves a lot of straight line connections between various nodes and structures within the brain.

In actuality, of course, these pathways of communication are much more complicated and convoluted. What you see here are the pathways of communication throughout the brain. These are white matter pathways, also known as the insula or the internal capsule of the brain by which information is distributed and communicated throughout the brain. It’s called white matter because these are myelinated axons. And so when you look at the tissue itself, it has a kind of a perilescent whitish sheen to it as opposed to what we call the gray matter, which are the unmyelinated axons of the outer layers, those six layers that were represented by those columns of the micro networks earlier. So for a given neural network, scientists have been able now to map out various important areas scattered throughout the brain that communicate with each other and they represent these as kind of nodes when you have a major communication aspect.

Default Mode Network (DMN)

One way of reflecting that is called a connectome. It is a comprehensive and detailed map of the neural connections for a given function. What parts of the brain connect to other parts of the brain in order for that function to be achieved and what is the strength of those connections? How much information is being sent? Now the easiest, I think, to digest representation is the one that you saw just a minute ago, which is those kind of straight line connections. This where regions of the brain that are involved in a particular neural network are represented in terms of the strength of their connectedness are in terms of the amount of information that is being shared within those areas. I want to run through a handful of identified neural networks in the human brain by way of example, and we’re going to start here with this one, which is the default mode network.

Now, the default mode network represents within the brain areas that are active when you’re not attending to anything in particular. You’re not attending to anything focused on anything external. You are awake, but at wakeful rest. You can think of this as the parts of your brain that are active and communicating when you’re daydreaming. Something my grandmother used to refer to as you’re just wool gathering. I don’t really know what that means, but-

Brad Dexter: Something new with sheep.

Jeff Snell: We didn’t have any sheep. My grandmother had no sheep. I don’t know where that comes from, but that is a term that is sometimes used for daydreaming. Your mind is just kind of aimlessly wandering. You’re not really focused on anything in particular. Now, considering that the default network is associated with the brain at rest, it’s quite widespread and the various structures and nodes within the brain and the strength and connectivity of those various regions can vary substantially. I’ve put for each of the brain areas, these neural networks that we’re talking about, the major parts and anatomical pieces of the brain that are involved and I have embold those that are kind of primary to a given neural network and those that are more secondary are listed in unbolded text there to the right. You will notice on just about all of these that aspects of the thalamus and the cerebellum are also involved.

Salience Network

Again, if you attended or have seen previously the cerebellum presentation that we’ve done, a lot of early brain research did not use any information regarding cerebellum and a lot of MRI scans, CT scans really mask out the cerebellum. And as a result, for a long time, we really didn’t know how involved the cerebellum was in everything, but you’ll notice in most of these neural networks, the cerebellum is going to be a piece of that as well. Now, one thing to go back here too again represents what is happening in your brain when you are not focused in particular. This network tends to correlate negatively with activity and task positive networks. So the salience network, for example, is involved in detecting and filtering important information, any kind of stimuli, external environmental or internal perceptions, a noise, a moving object, your stomach growling, pain with movement, our salience network acts as a controller.

It directs our cognitive resources between other networks when there’s something to which we should attend. So as an example, if our default network mode is our current mode, if we are just kind of sitting and daydreaming and then suddenly there’s a knock at the door, our salience network hears that, perceives that, detects it, and redirects brain resources to other networks to process and decide what should we do as a result of this new and surprising stimulation that’s come to us. What might be warranted to come to our attention? As you can see by the scatter of the primary nodes and activations, this particular network is compromised of a number of different identified nodes and primarily driven within the frontal lobes, the interior insulin and midbrain structures. It monitors constantly. It filters so that if it is not something we need to pay attention to, it does not bring that to our attention, but it is detecting and then switching other networks as a result.

Attention Networks: Dorsal Vs. Ventral

So these are the major structures that are involved. It initiates an evaluation of information coming in and prioritizes what should we attend to first. When the salience network becomes active through stimulus detective, the salience network signals the default network to deactivate and the dorsal attention network and central executive networks to come online to start working. Now there’s actually two distinct attention networks within the brain. There’s a ventral system which is involved in unexpected stimuli and the dorsal attention system, which manages planned and voluntary attention. These primary components are listed in the lower left and the subcortical and secondary are listed in the lower right on this slide. The attention system is anatomically distinct from processing systems that handle incoming stimuli that make decisions or that produce outputs. It is really distinct dorsal and ventral systems that operate like a seesaw. When one is active or heightened, the other is automatically suppressed.

The dorsal attention network focuses attention and is very closely tied to our visual system. It’s been compared to a cognitive spotlight that focuses and sustains attention on something that we are consciously and purposefully attending to. Right now with me looking at my notes, I am consciously attending to the presentation and what I need to say next, putting the next words in place. That involves my dorsal attention system represented in that kind of bluish color. You’ll notice as well from the standpoint of the ventral system, it’s a little stronger on the right side than it is on the left. So it’s a little bit more of a right hemisphere function. So you will imagine that damage to a right hemisphere or stroke in the right hemisphere is going to more impact that ventral system, that unexpected alert system that is there. The dorsal attention network’s considered a top down system.

So when we talk about top down, we talk about going from complex and funneling down into a specific point or action. So this is one of those. Motor system is also a top down system from the standpoint of the motor strip. It directs and sustains voluntary attention toward a goal directed stimulus. This system’s active when you’re intentionally directing your attention toward something that you’ve made a conscious decision to attend to. As you see here in the image on the left, this diffusing and unexploded ordinance. If that is your job, that’s probably the height of the utilization of the dorsal attention network. It is intentional and intense concentration on your every move. Now the ventral attention network, on the other hand, directs our attention when we have an unexpected stimulus that suddenly gets your attention. It’s when something you’re not attending to enters your consciousness. It doesn’t have to be as extreme as what’s pictured here or what is about to occur although that would likely activate a number of our neural networks, including our sympathetic nervous system and possibly some physical action directed toward the offender in this particular case.

It’s the ventral attention network that brings to your attention objects, sounds internal or external experience that are not your primary conscious focus and brings them to your attention. Now it is similar in that respect to the salience network, but the salience network is more involved with evaluating the importance of a stimulus, whereas the ventral attention network’s function is to reorient our attention to that unexpected stimulus. That can be internal or external as subtle as realizing I’m getting a little hungry or I feel like a headache might be coming on. That would be a somewhat subtle experience. If you’ve ever had one of your children leave a Lego on the floor that you’ve stepped on walking across the room, you will know that that is a much more striking and immediate and that is also your ventral attention network bringing that to your attention. And again, it pulls your attention immediately away from whatever you are consciously walking across the room for at that moment.

Central Executive Network

That now takes precedence. So the ventral attention network has the capability of, again, when it goes active, the dorsal system shuts down. So it crosses over there and those two seesaw back and forth. The central executive network, also known as the frontal parietal network, is responsible for conscious, focused, and active decision making. It’s all of the things that we typically talk about as being kind of your frontal systems, your bifrontal executive systems. But if you notice from the highlighting of the different nodes involved, it is parietal as well. It’s also midbrain and it is also cerebellar, although that’s not as represented within that particular graphic. It is rule-based decision making, but also integrating information from other brain networks that allows for flexibility. It is executing the results of that decision making and processing information. It’s responsible to initiate. It’s also responsible to inhibit. So it has some back and forth there as well.

Sensorimotor Network

The goals and goal oriented processes that our frontal executive system drives us toward can be responsible for internal or external processes, even when you’re just thinking about things and utilizing this particular network for making decisions or making choices. The executive network incorporates information coming in from our physical sensations, from our internal thoughts, from our prior knowledge, from our experiences, our raw emotions. All of this factors into how the frontal executive processes and this particular executive network utilize that information, combine it and make decisions and modify our behavior and actions as a result of those decisions. Now, lastly, I want to talk briefly here about the sensory motor network, which was actually one of the first neural networks that was identified. It’s often simply abbreviated as SMN or sensory motor network and it’s largely made up of regions of the pre-central gyrus and the post-central gyrus.

Remember I talked earlier about the central sulcus that goes kind of ear to ear across on your brain. So the bump on the front is the precentral gyrus and the bump behind it is the post-central gyrus. The precentral is motor, that’s all the motor signals being sent out. The post-central is sensory. That is all the information coming from your body, your internal sensory information, as well as external information that is going to that aspect. The sensoromotor information comes from our proprioceptive senses, our cutaneous receptors, our vestibular sensors are heavily involved with that. And then all that information is used to regulate a top down motor command from the motor strip to send fast and accurate movements. Again, that gets intercepted and mediated and modified by the cerebellum as well, heavily involved in all of that. At its most basic, this network is responsible for all of the sensory information that our brain receives and all the motor planning execution that gets sent out to our body.

Those network systems are closely tied with cognitive networks from a planning and goal directed behavior perspective and from an attention and executive processing process with emotional processing involved with that as well. And that can’t be underestimated the degree to which our emotional subsystems within our midbrain can impact these networks that we’re talking about. All of this is necessary to fine tune our behavioral response. Now there are numerous subsystems within this kind of umbrella network that have been identified, many of which integrate incoming and outgoing signals to other networks that we’ve been talking about. Again, cognitive and emotional processes. Disconnects within the sensory motor network have been linked to attention disruption as well as several different psychiatric disorders. And as Brad’s going to be talking about, emotional reactivity at a neural network level manifests in many different ways, which can, through neuroplasticity, become strengthened in pathways of communication with negative consequences.

They manifest as some of the things that Brad is going to be talking about. So at this point, I’m going to turn the helm over to Brad to review how these neural network disruptions can manifest in negative patterns of activity and even more important, the appropriate clinical response when those cases are identified.

Brad Dexter: Excellent. As Jeff hands the micro … Or not the microphone. Well, yeah, the microphone and the mouse over to me so I can drive. Jeff, can you just talk real briefly about how long have we known about neural networks? I mean, you mentioned the sensory motor network was one of the first ones to be discovered or discussed at length and how much is just the advancements in technology that we have, how much does that tie into our own understanding of neural networks at this point?

Jeff Snell: Well, I mean, really when neuro network came out was, as you allude to, when the technology allowed the manner of looking at what’s lighting up in the brain to do it. So when things like PET scans, SPECT scans, diffusion tensor imagery, when these became more refined over time, that’s when you can kind of really see, okay, when this action occurs or when we ask the brain to do something, these are the areas that are lighting up and these are actually the tracks and pathways in between these areas that are important to carrying that communication. So it matches up, it aligns perfectly with when medical imaging started bringing that to the table. So I would say late 60s, early 70s, some of the most basic work was done in this area. It continues to be refined, especially from the standpoint of integration of the cerebellum and where those pathways are, because that’s really only within the last 15, 20 years that that has become much more of a topic.

Why Neural Networks Matter Clinically

Brad Dexter: So I think about simple things that for you guys as an audience, you’ve probably been exposed to, whether it’s exoskeletons, artificial intelligence is a thing that’s on the top of everyone’s mind right now. Those things didn’t just pop up in the last couple years. They’re not just new advancements. They’ve been around for decades. It’s just hitting the mainstream currently. And so networks, neural networks have been around for a long time. They’ve been discussed. As you can tell, we can talk about them in a really dense manner, right? My job is to also talk about how do we bring this down to a clinical level a little bit more and what’s the utilization of neural networks? Why should we care about them?

Jeff Snell: We’ve talked about the weeds. Let’s talk about so what?

Brad Dexter: Absolutely. Absolutely. So just quick recap. The sensory motor network coordinates movement and body sensation. The default mode network is what’s running when we’re at rest. If we’re just ruminating on things, thinking about things or self-referencing, it’s that inward facing network. The salience network is kind of like the air traffic controller. So it’s monitoring, it’s filtering, it’s detecting and it’s switching as Jeff put it. The attention network has those two pieces. The dorsal side is top down, it’s intentional focus and the ventral side is the involuntary what the heck just happened capture. So the picture of the guy diffusing the bomb and the friend in the background with the paper bag.

Persistent Post-Concussive Syndrome

The central executive network is the working memory and task switching engine. It’s a network that lets you hold information to mind and use it actively. So just keep holding those in your mind for the next 15, 20 minutes here because I want to take you through three what could be considered frustrating diagnoses and we’re going to talk about each of these networks within that context. So we’re going to talk about functional neurological disorder. We’re going to talk about chronic pain and we’re going to talk about persistent post-concussive syndrome. I’m sure for you guys, for adjusters, clinicians, patients, we can all be frustrated with those type of diagnoses. They don’t fit cleanly on imaging. They don’t follow predictable healing timelines and they often get labeled as psychological exaggerated or just non-organic. I want to push back on that a little bit today. What I want you to take away from this section that I’m going through is that these conditions that people are living with have real neurological signatures, right?

They’re just signatures of network dysfunction rather than focalized injury. So it’s more difficult to show up on imaging. Once you map them onto networks, like Jeff has talked about, it gets a little bit more clear. So we’ll talk about why these are challenging in the first place. So let me just frame out that these are the cases that are often coming back to your desk over and over again. There’s really three things that make them hard. First, the symptoms don’t map onto a single injury region. It takes, well, let’s just consider a real patient picture for a moment weakness in the left leg, brain fog, light sensitivity, anxiety, all from the same event. You try to pin that on one anatomical structure and you’re going to fail. But if you map it onto networks, it makes more immediate sense. So the leg weakness involves a sensory motor network.

The brain fog is a default mode and central executive issue. The light sensitivity is the salience network’s filtering job breaking down and the anxiety reflects the default mode and salience networks getting locked into the wrong pattern. So in this case, there’s four different symptoms using four different networks. Same patient, same event. Secondly, standard imaging is often going to be unremarkable with these patients. The CT will come back negative, MRI is going to come back negative, yet the patient is clearly impaired in some way, shape or form. And this is where the system tends to break down for people because for you guys as case managers, for adjusters, even some providers are going to look at a clean scan and just conclude that there’s nothing wrong, right? So we have to be careful with that conclusion. A clean structural scan doesn’t mean that the networks are functioning normally.

The call out on the right side of the slide really makes a point. A CT or an MRI shows you the wiring of the house. It doesn’t tell you whether the lights are actually turning on when you flip the switch. So network dysfunction is about whether the signals are actually getting through, not whether the wires are technically intact. Functional MRI, diffusion tensor imaging can sometimes catch these things, but those aren’t standard reads in most work comp or even auto cases, right? Third, these conditions can develop after other catastrophic injuries. So at QLI, we’re dealing with a lot of catastrophic injuries. Someone has an orthopedic injury, multiple orthopedic injuries, spinal cord injury, even a stroke and weeks or months later, they can develop a chronic pain syndrome or even functional neurological symptoms on top of those things. The original injury here is going to be documented and accepted, but the secondary network condition often won’t be and that’s where those patients can fall through the crack.

So with that backdrop, we’re going to look at these three big complex diagnoses. We’ll look at FND first. So the definition that you see on the screen here, and I just want to read this carefully because the wording matters, these are genuine motor and sensory symptoms, weakness, tremor, gait disturbance, numbness, non-epileptic seizures. They’re real, they’re observable. You can see them, right? They show up on exam. The patient’s not just generating them on purpose. What makes them functional is that they’re not explained by structural disease. There’s no stroke, there’s no tumor, no demyelination on the MRI that accounts for what we’re seeing. Now for decades, this category got called conversion disorder or even psychogenic and patients would get dismissed. The implication was always, if you can’t se it on a scan, it’s got to be in your head and you have to be doing it on purpose.

That framing has done enormous damage. What we now understand and the network model that Jeff talked about is really central to this is that FND has a real neurological signature. Specifically, it’s a breakdown in communication between three of the networks that Jeff covered, the sensory motor network, the salience network, and the attention network. So let me just give you a concrete example, functional tremor, say in the right hand. Here’s what’s happening at the network level. The sensory motor network is generating movement, but the salience network is actually misreading that as threatening or alien, right? It’s not their own. And the dorsal attention network, which would normally be able to override and consciously stop the movement is somehow disconnected from that loop. So the patient is watching their own hand tremor. They know that it’s their hand, they can’t will it to stop and it feels like they’re losing control.

And actually the more that they try to stop it, the worse it can get. That experience in and of itself can be pretty terrifying to the patient and it’s also exactly what the network mismatch would predict. So here’s the clinical tell that we’ll look at with functional neurological disorder and it’s pretty interesting. A functional tremor is often going to decrease when the patient is distracted. So you ask them to do a complex tapping pattern with their other hand and the tremor frequently quiets. That’s the dorsal attention network finally getting hooked back in. It’s the top down control reasserting itself in the language that Jeff used. It’s not malingering. Malingering would do the opposite. That’s more of a network signal that’s giving us a clue. The call out on the right here in the purple box is the part that I want you guys to really internalize.

So FND is one of the most common reasons that people get referred to neurology. It’s not faking, it’s not a psychiatric diagnosis of exclusion and it’s absolutely not something that the patient can just snap out of. What it does respond to is a coordinated team, PT, OT, sometimes speech, neuropsych, neurology, all working together with a patient who’s been told clearly this is real and we have a path forward. So the framing of the diagnosis itself is actually therapeutic. The way that it’s talked about as a team and with the patient is actually therapeutic. So if you guys take nothing else away from this slide, when you see the F&D label come across your desk, it’s not a red flag for fraud. It’s really just a signal that the patient needs a coordinated network informed rehab team and they need it sooner rather than later. All right.

Chronic Pain Through A Network Lens

Chronic pain. This might be one of the more clean examples of why the network model matters in rehab. So we’ll take a little bit of a moment here. So just look at the comparison on the screen from left to right. Acute pain is what we all understand pretty intuitively, right? So you sprain an ankle, the tissue’s damaged, pain is going to be generated as a signal that says, “Hey, stop using this and just let it heal.” It’s localized to a specific area. It’s more predictable. It tracks with the tissue injury and as the tissue heals, the pain starts to resolve, right? Standard analgesics work because they’re acting on the same pathway that’s generating the signal. Now with chronic pain, and this is where we’re talking about pain that’s persisting beyond about three months well past that point of tissue healing, it’s going to operate on a completely different mechanism.

This is where the network model really becomes crucial. When pain becomes chronic, three of the networks that Jeff walked through become heavily involved. First, the salience network. So remember just four words, for the salience network, we’re monitoring, filtering, detecting, and switching. In chronic pain, all of those go sideways. The monitoring is stuck on, the filtering is broken, the detecting is firing for things that aren’t actually threats and the switching never lets the brain move on. So a patient with chronic low back pain is going to tell you that a light brush of fabric on their back hurts, that walking on a slight incline now hurts, that movements they used to do without thinking now hurt. It’s not exaggeration. It’s the salience network labeling normal sensation as threat when it shouldn’t be. Now, I know we’ve done a number of chronic pain webinars at QLI and that language is not any different.

It’s just putting networks and mapping those directly onto the same language that we’ve typically used. Second, the default mode network. So the DMN, as Jeff talked about, is what fires when we’re at rest, when we’re just ruminating on things or replaying events in our minds. In chronic pain, the default mode network locks on to the pain experience. The patient lies in bed thinking about their back. They’re in the car thinking about their back. They wake up at 30 AM thinking about their back. The thinking itself is reinforcing the network firing pattern and it becomes self-sustaining. This is also why chronic pain is so closely tied to depression and anxiety. They share the default mode network as a home base. Third, this one is a little more subtle. The ventral attention network. So Jeff’s bomb on the road image. The VAN is involuntary. Something just grabbed my attention network.

In chronic pain, the VAN is being hijacked by pain signals dozens of times an hour. So every twinge, every shift in position is capturing attention involuntarily. The patient can’t focus on a conversation. They can’t focus on a task at work on a TV show because their VAN keeps yanking their attention back to their body and in this case the back. So this is why patients with chronic pain so often describe the pain as spreading. It’s not focal anymore. It’s everywhere or just different than it used to be. Again, it’s not exaggeration. That’s just the networks generalizing in the background. The right hand card here is the implication for care. If you keep chasing that original tissue site, so more injections, more imaging, more surgical opinions, you very rarely are going to resolve chronic pain because the tissue isn’t really the driver anymore, the networks are.

What does work again is a combined approach. Physical therapy that uses graded exposure to gradually retrain the salience network’s threat signal, neuroscience-based pain education so that the patient understands what’s happening in their own brain and psychological support. So often using cognitive behavioral therapy or acceptance commitment therapy to address the default mode network’s role. For case managers and adjusters, those of you on this webinar, the practical takeaway is really when you see chronic pain that isn’t responding to that standard musculoskeletal care. It’s not a sign that the patient’s being difficult. It’s more of a sign that the case needs a brain network informed team, not another orthopedic consult. All right, persistent post-concussive syndrome. This is the third complex case that I want to walk through. Here’s the puzzle with these cases. The vast majority of concussions are going to resolve within two to four weeks. The patient gets symptomatic care, gradually returns to activity, and then they’re back to baseline within a month.

But there’s a meaningful subset of folks where the symptoms are just going to persist for months, sometimes years. And critically, when these patients get re-imaged, the imaging looks fine. The structural injury is going to appear to have resolved and yet the patient is still very much symptomatic. You guys probably have this person on your mind right now. This is one of the cases where we can see a lot of frustration from adjusters, from primary care providers, from clinicians, because there’s an implicit assumption that says, “Hey, the scan’s clean, times passed, the patient should be better, but if they’re not better, something else has to be going on here. Maybe there’s litigation, maybe there’s secondary gain, maybe it’s psychological. Let’s just push back on that for a moment using the network model because it gives us a language to do it. If we look at the symptom card on the right hand side of this slide, we’ll just walk through how each one of those persistent symptoms can map directly onto the networks that we covered earlier in the presentation.

So headache, that is the salience network firing on high alert. Again, just for words, monitoring, filtering, detecting, switching. Post-concussion, the salience network’s filtering job breaks down. So the normal vascular and muscular sensation that should be filtered out of awareness gets labeled as threatening and the brain experiences it as headache. Brain fog and memory complaints. These are going to attract to the default mode the central executive and the attention networks all working against each other. They’re fighting. The central executive network can’t sustain working memory. The dorsal attention network can’t hold focus where the patient wants it and the default mode is intruding when it shouldn’t. The patient is sitting down to read an email and they can’t hold the second sentence in mind by the time they finish the third. That’s a central executive network failing. They start the email, then find themselves thinking about something unrelated for two minutes.

That’s the default mode, hijacking the attention network. Not laziness, it’s three networks failing to coordinate with one another. Light sensitivity. This is pure salience network filtering failure. So the normal ambient light is being labeled as a threat level stimulus. The salience network is supposed to filter most light intensity out of awareness, but in persistent post-concussion syndrome, it can’t. Sleep disturbance. This is involving the default mode network and the cerebellum. So Jeff made the point earlier that the cerebellum is the underappreciated hub and one of its jobs is helping to coordinate the rhythms that allow us to transition between wake and sleep. When that’s disrupted, the default mode can’t quiet down at night and patients will describe their brain as racing when they’re trying to sleep. Fatigue. This is what happens when networks have to work harder to do their normal job. So imagine running a router at twice the normal CPU just to push the same amount of data through.

The patient is exhausted by mid-afternoon because their brain has been over recruiting networks all morning just to do the tasks that used to be automatic. Irritability, default mode and salience network interaction with emotional regulation. The threshold for emotional response is going to drop because the salience network is overcalling threats that the default mode is ruminating on and the default mode is ruminating on them. Dizziness, this is a sensory motor network and cerebellum problem again. So the cerebellum’s critical to spatial orientation and as Jeff mentioned earlier that it gets routinely underread on standard scans. So the vestibular cerebellar miscommunication ultimately produces the classic post-concussive dizziness that will get treated for these cases. So when these symptoms are persisting, they’re not vague complaints. They are confusing to us, but let’s be honest, as we go through these neural networks, they’re confusing as we talk about them. They’re something that we’re still learning, but every symptom on this card has a fingerprint.

Network-Based Rehabilitation Strategies

And if you start to understand these networks a little bit better, you can start to see those fingerprints better too. So the line at the bottom of the card here is the line that I want you to walk away with. These symptoms are real. They’re measurable with the right tools and they’re network based. The treatment implication, again, is multidisciplinary here as well. Okay. I wanted to put a synopsis on this slide just to kind of show how you treat each one of these networks. This is kind of the thing that you should walk away with and I just want to go through the table here. So in the sensory motor network, when this misfires, you’re going to see movement that’s out of sync with intention. The patient wants to move, but they can’t, or they move abnormally despite trying. How do you target that? Well, PT can target it with motor reeducation at the core.

Task specific retraining, like practicing the exact functional movements that the patient needs, not just generic exercises. And then vestibular therapy when balance and spatial orientation are involved. So just notice that this row is going to cover a lot of the FND presentations and the dizziness piece of persistent post-concussion syndrome. In the salience network, we covered those four words that Jeff hit on several times. When those misfire, the filter is going to fail and ordinary input gets flagged as threatening. That’s a patient who can’t tolerate light. They can’t tolerate sound. Their chronic back pain spreads into territory that was never hurt. What targets that? Graded exposure, gradually reintroducing the input that the brain has decided is dangerous. In small enough doses that the salience network can actually update its threat signal. So pain neuroscience education is also a way to treat this. Teaching the patient what’s happening in their own brain, which itself reduces threat perception.

Visual and sensory therapy is another thing and mindfulness-based interventions, which have really, really good evidence for retuning that salience network. The default mode network, when it locks into rumination, the patient can’t get out of their own head. What targets that? CBT. So cognitive behavioral therapy and acceptance and commitment therapy. Those are the workhorses in this area. Sleep optimization matters because the default mode network doesn’t get its normal off cycle without proper sleep. And then behavioral activation, getting the patient engaged in meaningful activities is going to pull the DMN out of its loop. The attention network, both the dorsal top down piece and the ventral involuntary capture piece. When this misfires, patients can’t hold focus and they get pulled away by every internal and external distraction. How do you target this? Cognitive rehabilitation done by speech, language pathology or OT, direct attention training. And this is the part that’s really underrated pacing and energy strategies because attention recovers better with structured rest and with brute force effort.

And we hear that all the time from patients, especially these high driven ones, they’re trying to do too much too fast. Central executive network is the last one on the slide here. So the working memory and task switching network. When this misfires, the patient sits down to read something, read an email, and they lose the thread by sentence three. How do you target that? Again, speech language therapy is going to be huge here. Cognitive rehab, compensatory strategies, helping with calendars, lists, checklists, phone reminders. Those things don’t fix the network, but unless the patient function while the network is retuning itself. And then the line here at the bottom of the slide is really the takeaway for you guys. When you’re authorizing a rehab plan, the question isn’t, did we order the standard protocol for whatever this case would be? The question might be, did we look at this specific patient’s network signature?

Key Takeaways & Interdisciplinary Care

Did we authorize the therapies that target the networks that are actually misfiring for them? So that’s the difference really between check in a box and building an effective plan. All right. I am one minute over, but I’m going to just hit this last line on this slide. You guys should be thinking about individualized care informed by this network model by neural networks and that should be the standard, not the exception. We saw kind of laced throughout here the need for multidisciplinary teams having several different types of clinicians looking at the main issue. Okay. I can touch on these key takeaways, but I know…

Steve Kerschke: We’re… If this question’s poorly timed, tell me. You’re talking about interdisciplinary care and a comprehensive team. If I’m thinking about this in terms of the care continuum at what level, like is this a post-acute program? Is this an outpatient program? Describe where I’m looking for that and where do I go?

Brad Dexter: Yeah. So I think in the cases that we’re talking about today, Jeff, feel free to chime in on this. These are not things that show up right away, right? These are things that show up with time, but…

Jeff Snell: This is a maladaptive adaptation of our brain.

Brad Dexter: Absolutely. So we talked about catastrophic injuries and oftentimes you might have chronic pain or an FND that shows up down the line, but we tend to get confused by what’s going on. This doesn’t map with the injury to the region in the brain. Why are we seeing what we’re seeing? That should just be a cue that, oh, there’s something else going on with the networks perhaps and we need to address it there. So where should this be addressed in the continuum? Oftentimes it’s going to be a little bit later in the continuum of care.

Steve Kerschke: Yeah. And the other thing I think I will add that I notice is sometimes we’ll get a referral and either the referring source will say like, “I’m not really sure who to refer them to. Do I need PT, OT, speech, or psych?” Or we’ll get a single referral for, let’s just say PT and we’ll start with that and we kind of layer things on. So I think that there’s a lot of ways to look at this to keep utilization and check and sometimes depending on where you go and what clinic and the understanding of the clinician, a PT can handle a lot that is very versatile in some ways. I have a person who’s, she’s not at post-concussive yet. We got her when she’s three, four weeks out from her injury, which there’s probably a lot of spontaneous recovery, but just in my PT session we were talking about pacing. We were talking about some of her cognitive load. We were talking about her sleep, many things that would maybe land with an OT or an SLP that within my session we were able to navigate. I think it’s important not to be super black and white about this. It’s really just, can you find the right providers who understand the networks and other aspects of care because a lot of times, especially if caught early, you can do a lot in a very short period of time.

Brad Dexter: Yeah. So let me just touch on the few things at the bottom of the slide too, because I think you hit on the build teams part, Steve, but I also want to touch on the stay curious as the evidence evolves. I mean, we have 20 years of more clear evidence on some of these things and that the neural networks go back further than that even. But one of the articles at the end of this has, it’s like, what have we learned in the last 20 years of studying these networks? And my point in saying this is what we know right now and what we’re talking about right now is going to be different 10 years down the line, 20 years down the line as technology continues to advance. Absolutely. We certainly have to continue to understand brain anatomy at a regional level. We have to understand, but we also need to more continuously evolve our understanding of neural networks and the roles that they play.

We all know that the brain is complicated. We all know that it’s complex, right? We all don’t have to understand it at the deepest level, but recognizing that neural networks are a real thing and they’re impacting our patients even when we can’t see it on imaging, I think matters in the way that we treat…

Jeff Snell: And we’re talking about the brain in isolation, which is not separate from the body.

Brad Dexter: Yes.

Jeff Snell: So all of that even makes it more complicated.

Brad Dexter: Yes, absolutely. And Steve, you hit on the build teams part, like maybe you start out with on provider, but as you start to recognize, and I think so many of you on this webinar do this pretty well at a high level already. So maybe this is just reinforcing your own decision making, but you recognize that your patients aren’t going to receive the best care if it’s just a siloed approach and you need to build a solid team of care providers around them for them to get the most excellent care.

Audience Q&A

Steve Kerschke: All right, let’s do the poll question. Now is a great time for our audience to ask questions if you have them. Let me get the poll question up here.

Brad Dexter: We made the poll question really, really difficult, didn’t we?

Steve Kerschke: Yeah. It’s always a stumper.

Jeff Snell: Yeah.

Steve Kerschke: This is another really tough one

Jeff Snell: Talking about all … Yeah, go ahead.

Steve Kerschke: No, I mean, I was just talking about the complexity

Jeff Snell: Talking about this, just in the process how your mind on a separate track can operate. While we’ve been talking about this today, I’ve been thinking about neural networks from the standpoint of thinking back as a parent with our kids and when somebody would come over who didn’t have kids and go, “How in the world can y’all stand all this chaos and this noise?” And it’s like, well, our brain has adapted to that and that has become our new normal. Those things are not catching our attention. They are not pulling us away from our conversation or anything because it’s now our brain has adapted to that. The other interesting thing as a parent that comes up is when all of a sudden you realize it’s too quiet and that’s when your ventral attention network goes, “Hey, what’s happened?” And that sends it to your central executive network to go, “Okay, what do I need to do about this?” And it puts you into action. So yeah, these networks, they’re not exclusive to brain injury. We’re using them every day and in every way.

Brad Dexter: And I think as a physical therapist, it’s really helpful for me to think about the networks because I can see what’s being manifested from a symptoms perspective, from a movement perspective in a patient and it helps me understand how I can drive neuroplasticity even better. So it helps me focus where my treatment needs to be in order to impact a certain system. It also helps me see and reinforce from an educational standpoint with the patient why perhaps interdisciplinary care is beneficial for me.

Jeff Snell: Yeah. Why focusing on one particular aspect to the exclusion of these other things that also influence that particular manifestation is not going to be effective.

Steve Kerschke: Just to reiterate the education, you all said it multiple times, but I think in practice when people understand generally what’s going on and why they’re feeling what they’re feeling, it’s so helpful, right? I mean, this might be an unfair question for you, Brad, but when you see, you gave the example of a movement, call it a movement disorder or a movement presentation where the right arm wasn’t behaving the way it should, but by tapping in and using the left arm, you were able to tap into a different network. Would you say that that’s a good place to start in terms of treatment?

Jeff Snell: Tap into a network. I see what you’re doing.

Brad Dexter: Yeah, very good. Very good. Yeah. I mean, so the concept there was distraction with functional neurological disorder and if the patient themselves in that circumstance is continuously kind of trying to focus and change what’s happening and distracting them to tap in or rewire that network in some way, it’s a similar concept as pacing or energy conservation strategies that we talked about for persistent post-concussion syndrome. What you’re actually doing is you’re helping to rewire that network in some ways. You’re taking away attention from the thing that’s going haywire and putting the attention on something that is not.

Jeff Snell: Yeah, and the same way with chronic pain, you don’t want to ask somebody to rate their pain 50 times a day because you are reinforcing that paying attention closely to something that you want to pull away from.

Steve Kerschke: All right. We have one minute, but we have one question that I want to make sure we hit. The question is with FND with someone who’s had it greater than 10 years, do you have thoughts on how to integrate against a surgeon who wants to do physical interventions that have failed three times previously? So my take is you still have potentially a surgeon or a medical player who is taking a more medical approach.

Jeff Snell: Isn’t there a saying about doing something repeatedly and expecting a different

Steve Kerschke: Outcome?

Jeff Snell: I think so. Okay.

Steve Kerschke: I think that my take is that, and we see this a lot too, is when there’s a medical provider on board with a significant amount of power, usually it is a physician. It’s a real challenge and I think I would encourage some sort of consideration for maybe a second opinion if possible. I know there’s constraints related to work comp on that. Some sort of education with the surgeon. Again, constraints with that. I don’t know that there’s a magic answer to this one, but I don’t know if you guys have thoughts on that.

Brad Dexter: I mean, Jeff, feel free to chime in and there’s a lot of information that I, a lot of context I don’t have on this case either, but I’d be curious who the other providers, to put it the way that you put it, that maybe have a little bit of power or say in that. If there are any different perspectives that could be given to the surgeon, is the surgeon the one that’s driving a lot of that care? Has a neurologist been seeing a functional movement specialist? I’m sure if this has been happening for 10 years that they’ve gone through some of those things. What amount of education and buy-in do they have in regards to their functional neurological symptom diagnosis?

Jeff Snell: And the degree to which the whole system is being considered. Again, when you try to treat one symptom in isolation, that typically does not work very well with human beings because we’re very complicated.

Brad Dexter: Yeah. I don’t know what type of surgery we’re talking about either.

Steve Kerschke: I think it was a spinal cord stimulator, it sounded like. I mean, I think at the end of the day, that’s a tough one and we have found it to be quite hard to connect with other providers, typically physicians, because they’re busy. They don’t have time for just a random conference call. But our team has been in lots of situations, not necessarily exactly like that, but if there’s a physician who’s not on board as a team, we have tried to reach out, tried to collaborate, tried to send notes. We’ve done a lot of different things, but unfortunately that part’s pretty complex.

Jeff Snell: Yeah. You ideally want a specialist who also has the capability of looking holistically at the case, not just their specialty.

Steve Kerschke: Yep. All right. We are a couple minutes over. Want to be respectful of everyone’s time. Jeff, Brad, thank you so much for the presentation.

Jeff Snell: Brad, thank you so much. That was fun.

Steve Kerschke: Yeah. We’re always glad to have both of you and such great information. I know I always learn something just listening to you guys go back and forth. So to our…

Webinar