Dr. Kim Bullock  

Welcome back to Psychiatry XR, where we have worldwide conversations about the use of extended reality in psychiatric care. I'm your host for this episode, Kim Bullock, and I'm joined by my spectacular co-host, Jessica Hagen. Hi, Jessica.

Jessica Hagen  
Hello, everybody.  

Dr. Kim Bullock 

in this episode today, we're going to be joined by someone else who's pretty spectacular and also a Stanford faculty colleague of my own department, Dr. Hadi Hosseini, and Dr Hosseini is an associate professor in the incredibly competitive research line in the Psychiatry and Behavioral Sciences and the director of computational brain research and intervention Lab C-Brain. His research focuses on integrating multimodal neuroimaging psychological testing and computational and engineering approaches to elucidate signature changes in brain structure and function in neurological and psychiatric conditions. One of his areas of research includes developing and testing cost-effective and neurobiologically valid Virtual Reality assessment tools for early detection of Alzheimer's disease. Thank you so much for joining us, Dr. Hosseini.

Dr. Hadi Hosseini

Great to be here, Kim, and good to see you guys.

Dr. Kim Bullock  

Yeah, it's great to have this chance to have a conversation in depth about your work. And so I know we both kind of got interested in VR around the same time. I don't even remember how we met, whether it was just around the water cooler or something and talking about VR, but I think it was almost a decade ago, and I was wondering if you could tell us your story about how you became interested in integrating virtual reality into your work. I don't think I really know exactly the whole story myself. 
Dr. Hadi Hosseini 

Yeah, that's a great question and I think it was definitely more than a decade ago, and when I joined the Department of Psychiatry, I realized a lot of the assessment tools that we use, particularly for like cognitive assessment, are really not ecologically valid. They are pretty good in terms of picking up very drastic, for example, differences in cognition. But then they're not sensitive enough to pick up very subtle differences in cognition. And then also, a lot of them are not necessarily biologically valid, so there's a kind of like gap between behavioral manifestation or cognitive outcomes, and then what's going on at the biological brain level? And I think VR actually at that time, well, there was, like, certain advances in technologies, right? So it's been around for many decades, but then I think a decade ago, it was good enough to test it in kind of like, you know, affected population and psychiatry, the quality of the visuals were improving and so on. And that's how, actually, I got interested in seeing how we can use VR for assessing some of the cognitive functions that cannot be tested with paper and pencil or computerized tests.

Dr. Kim Bullock

Got it, it’s really just like an enhancement on what we have, just to be more accurate and valid it sounds like…

Dr. Hadi Hosseini 

Correct. Yeah. And then I think I started to see how we can test it, for example, how we can use this classroom for attention testing, right where, for example, kids with ADHD would get these realistic distractors that may go on in an actual classroom, for example, or assessing video spatial or navigation skills in older adults, where it's often pretty difficult actually to test with just simple computerized tests.

Dr. Kim Bullock 

So how did you develop that interest and actually be able to integrate it? Did you just reach out to networks of people? Or how did you logistically get savvy with the technology?

Dr. Hadi Hosseini 

I think being part of, of course, SPIT-C, connecting with yourself and Jeremy Bailenson, and then also even, I think at that time, I was talking with Skip Rizzo at USC at some point, and trying to see how we can get this going. I think that helped a lot. And then at the same time, I think using some of the resources that became available through SPIT-C. But again, like my background was also in engineering, so it wasn't a tough task to kind of like integrate this VR with my other programs, maybe for the audience. I do a lot of, kind of, like multimodal neuroimaging and computational techniques and use those in my research and adding like VR to that was not challenging. I would say, though, at the beginning, let's say, trying this, for example, on older adults with the older technology, it was a little bit more difficult because of the cybersickness, you know, those kinds of, like problems that you had, again, like with the advances in these kind of like new VR goggles. I think that's becoming much and much less of a problem.

Jessica Hagen  

How did you actually make your way into studying early detection for Alzheimer's using virtual reality?

Dr. Hadi Hosseini 

That's an interesting question. So I started actually becoming interested in Daydream because I was very interested in kind of training, actually, like, how we can improve brain functioning, or networks using, I would call it, like, these organic ways of improving the brain. I was very interested in kind of like, seeing how we can combine different types of kinds of training to target a specific brain networks. Can we see improvement in brain structure and function, to looking at biological changes in response to that. That's how I got into kind of like running this clinical trial, actually, for looking at how a multi-domain kind of training long-term impact brain structure and function in older adults with and without memory problems. And at that time, I realized that also, again, like some of these assessments, tools that you use, are just so rudimentary, and then they're so prone to different types of noise, you know, cultural background, language, a lot of things. And that's how we actually got into the Alzheimer's field.

Dr. Kim Bullock  

Got it. That brings me to one of the urgent questions I was going to ask you, because you, as a neuroscientist, you are probably looking at different aspects of cognitive functioning and coming from it from a very different perspective. And I know a lot of us, clinicians and physicians and some of them may be listening, are probably wondering about this too. Ever since that kind of brain game craze about a decade or two ago, with companies like Luminosity making these false claims about training effects on memory enhancement, most of the neuropsychiatric disciplines have now become increasingly skeptical that cognitive training can make an impact on prognosis and neurodegenerative disorders. So even recent Cochrane Reviews that do these intensive systematic reviews came out finding only low-quality evidence suggesting any improvement, and these were usually just in a very short run. So yeah, love to hear what made you believe that cognitive training, because you were kind of coming into it at that point, cognitive training could make an impact.

Dr. Hadi Hosseini 

It's funny, yeah, it's been a controversial field for a while, and still is. And I think one of the problems that we have in the field of kind of training is that I think there was, at least at the beginning, there's been a lot of trials or studies that were not really well designed, like there are so many differences in the way that, in the protocols, duration of the training, types of training, the kind of domain that they are focusing on, let's say, just focus on dementia, for example. 

Well, and then there were so many different results, right? Like negative outcomes, positive outcomes. And to me, I think the effect sizes that you would see from these kinds of studies are generally low, so you expect to see so much variation in the outcomes. And I think that kind of like contributed to this controversy that okay, so these games, or training games, may not be helpful. 

And at the same time, also, there are some claims about these helping improve Alzheimer's disease and so on, like, some claims that are very general and can be interpreted, I would say differently. My understanding is that Alzheimer's disease, so it starts to develop maybe 10, 20 years before the clinical symptoms are kind of like appearing like so these amyloid plaques and tau tangles like, you know, start to build up decades before actually manifestation of the clinical symptoms. And I don't think cognitive training is gonna clear up the amyloids or tau tangles, right? So I think that's the kind of like the confusion. 

So disease is there with kind of training you are not going to remove amyloids or touch angles, so the pathology is there, keeps developing. I think the argument that I am for is that I think cognitive training can kind of like a scaffold right, or strengthen right the brain networks functionally to be able to withstand some of those pathology right? And then data showed that if you can, for example, delay the clinical onset of, for example, dementia for five years, the number of people who would develop clinical AD would reduce, like, 40% that was like several million people reduction in the numbers.

You know, a lot of these cases are sporadic. Usually happens after the age of 65, so five years in that age range is a lot. And I think that was my mindset that okay,  if we can really improve brain networks, if you can improve, like, your cognitive reserve, basically, you may be able to delay the onset of dementia for a few years, and that would basically be still a lot. I should actually mention this trial that I just briefly said, that I got into the film with that, like running that trial, and that was my first clinical trial that I ran, and it was interesting. 

So I looked at these previous studies, and I realized first, there weren't many studies that are like on MCI. There are a lot of studies on older adults aging, there weren't at that time many studies on MCI population. So MCI is a small kind of impairment. This is kind of like a pre-normal stage of Alzheimer's disease. They have a higher risk factor for developing AD. Second, a lot of the trials are very short, so two months, most of them, like, some of them could have been like three months. So I said, what if we do it for a longer term? 

So if the effect is not that huge, so we do it for six months, if we don't see anything, then probably it's hard to see anything, right? So that was another one. And then I also, at that time there weren’t many multi-domain cognitive training, so they were either focusing on memory, working memory, and so on, right? So, then I said, let's build a program combining all of these different cognitive domains trying to enhance whole and when we know that, like, okay, prefrontal cortex, for example. So memory function is kind of like observed by this temporal lobe, right? 

So these hippocampus and frontal cortex and so on. But then we know that, like a lot of the decoding, coding is happening like in the frontal cortex, and then you may need to, then, if you want to strengthen the prefrontal cortex, you want, for example, different types of tasks, right? So we kind of put together this multi-domain content training. 

So in this trial, actually, it had two groups. Both groups had MCI (Mild Cognitive Impairment), but one group did the crossword puzzle training as our control, and the other group did this multi-domain kind of training, focusing on memory, executive function and so on. And that was another, actually addition to previous studies that we said, okay, let's do this more active control group that if we see something, that means something, right? So if you just say, okay, I'm just looking at one group and I see changes. Those changes could be because of a lot of different things. 

But then when you try to have a control group with kind of like, you know, a lower level of cognitive intervention, which in this case, it was the crossword puzzle training. These people, like these older adults, actually did the assessments and also an MRI scan in the lab. And then they went home, we sent them these intervention protocols for crossword puzzles and for the client to train. 

And they did this for six months, and they came back, and then we scanned them again. There were some issues with the COVID, like, you know, we couldn't bring some of them back. But anyway, what we found was that the primary outcome, which was the fluid cognition, which is a combination of, again, memory, executive function and so on, we found actually significant improvement in cognitive out fluid in the cognitive training group compared with the crossword puzzle training group. But most importantly, we also looked at differences in brain data, brain structure and function. 

One of the interesting findings was that looking at these kinds of like white matter tracts, properties. These are older adults, so both groups actually were going down in terms of the integrity of these white matter track. But then what we found was that the kind of training group actually went down much less than the crossword puzzles group, which makes sense. So I don't believe that. Like, okay, so with kind of training six months, we're going to kind of, like, enhance the white matter but at least it seems that what the data suggested was that it at least slows down that process again. This was a narrowly clinical trial that needed to be tested in larger sample, I think. But then those were kind of pretty interesting, actually.

Dr. Kim Bullock  
Okay, well, you're making the skeptic in me a little bit more excited about this. But yeah, it sounds like the quality of the research, you've got a longer trial, better design, and you've got the neurobiological data as well, and you're kind of focusing on mitigating some of the decline and just prolonging the duration of the illness. That makes a lot of sense. I've been dying to ask you that question for a long time, so thank you so much for taking the time to explain that I noticed I got a flyer about your virtual reality memory study for older adults. Are you still recruiting for that?

Dr. Hadi Hosseini 

Yeah, thanks for mentioning that. I think that should have been the main part of this podcast. They should talk about going back to your first question about how I got into VR right? So again I think VR can help a lot in terms of providing more realistic right or ecologically valid assessment right or understanding behavior and testing or assessing cognition. And that's, yeah, how we kind of, like, started this study.

To give you a background, actually, so in older adults who have Alzheimer's disease, these basically, little tangles start to actually, like, build up in the medial temporal lobe and in this area, as they call, like, entrenched cortex, and then they spread to the hippocampus and then parietal core. 

Interestingly, these structures are pretty important for visual, spatial, and navigation skills. And then the question that we had was that maybe with using virtual reality, creating an immersive environment, having an ecologically valid assessment for assessing these visual, spatial and navigation skills, maybe we'd be able to pick up these very early changes in cognition, right, that may later build up and then become evident as a big memory problem.

And then we started to work with Anthony Wagner's group, specifically Tammy Tran who's a postdoc with him. And then Jeremy Bailenson, at the virtual reality lab at Stanford, who started to build up these assessment tools to specifically focus on navigation and visual spatial skills. Okay, so these are probably the earliest functions that may be impacted in early Alzheimer's. 

And to answer your question, we are done with that study, and we have some interesting preliminary results, relatively smaller studies, we had around like 25 people in each group. And by groups I mean one group who had mild content impairment, again, like epidermal to Alzheimer's, and then another group of healthy older adults who are clinically unimpaired and continually normal. 

So we compared this also with the standard neuropsych assessment. And let me tell you a little bit about these tests. So we had one test where a subject would go to this room, we call it Sam's room, and then there were different objects in different places, so it was mimicking, kind of like a living room of a participant, and the participants would interact with these objects for like, a certain period of time to learn the location of these objects. For example, the remote control is here, umbrella is over there, and so on. After this period of encoding where these objects are without telling them, right, you tested their memory. Like in the next phase where they entered the room, the objects were not there. We were asking them, okay, can you put these objects back into where you found them, like in the past, over the past, for example, 30 minutes of encoding. And so that was one of the tests. 

So one of the beauties of actually using VR was that so we can monitor the exact location of each object, right? So you can't really do it in the actual testing, right? So they would put the umbrella there, we could tell not only if the rough location was correct, but also how exactly they were in terms of spatial memory. At the same time, we are also recording the eye tracking data and trying to see okay, their gaze, for example, patterns and so on. 

Another task was that actually they would see an avatar in this virtual room that would go over different kind of, like blocks presented on the on the floor, and the participant observer, what's supposed to actually kind of like, encode that sequence or the path that the avatar would take, okay, memorize that path and then try to recreate that path like, okay, now it's your turn try to follow the path that answered it, and then some other also block design memory and so on. 
Anyways, so those were the tests, and then when they looked at the outcomes, actually, what was interesting is that for this first test, for example, we call it object location memory, right where they're in this room, trying to memorize the location of objects. So when you compare the accuracy in terms of like, how they place these objects were, I would say trending in terms of differentiating between MCI and cognitive normal controls. But when we looked at precision accuracy, we found that actually, older adults who didn't have memory problems had significantly better spatial memory precision compared with older adults, which was very interesting, and the effect size is pretty large. And again, like that's also adds to the you know, how VR can be helpful, right, in terms of picking up these subtle things, being able to monitor behavior from different aspects that cannot be really achieved, or if you want to achieve in a realistic environment, it's gonna like be a lot of effort.

We also found interesting results in terms of gaze pattern, the number of, for example, saccades that they would do following the avatar path, for example, that would actually differentiate between MCI and the control and also correlated with their memory problems. 

But then the last one, which is more important, was that this special memory precision recall in the anti-object location memory test was correlating with this structure of this tract, which we call it, single and hippocampal tract that connects hippocampus and middle temporal regions to the parietal regions. Okay, and so there was actually a huge association between those. And not only was it an association between performance and the VR, but also we could actually use VR performance across these tests to predict actually the quality of this track. 

Dr. Kim Bullock

Wow. 

Dr. Hadi Hosseini 

This was very fascinating. Again, like we identified a couple of measures. So we had around maybe 10, 20 measures, but we identified like, two or three measures that were the most distinguishing between MCI and controls, but also, kind of like predicting these brain structures. Wow, it was relatively small samples, like 25, 25 but the effect sizes that we found were really big for some of these VR measures, which we are hoping actually to replicate and validate, again, like, in a larger sense.

Dr. Kim Bullock 

Got it. Okay, yeah, it has so much potential for neurocognitive testing, the precision and the standardization and predictive value that these things think you're paving the way for that.

Jessica Hagen  

So you've obviously done studies just like you described, where you studied cognitive decline in individuals, right using these different tests. Have you done those same types of studies using virtual reality, and have you seen a difference between when you use virtual reality and when you don't?

Dr. Hadi Hosseini 

That's a good question. Actually, I should have mentioned that. So I think one of the purposes of this study was that, how much better or worse the VR measures are compared with these standard tests. And then we had different types of actual, standard tests. Some of them were very direct equivalent to some of the VR tests that we did, and some of them were just kind of like standard tests that our people are doing in the field. 

And then what we found was that comparing, for example, memory precision recall versus performance in these other tasks like ray auditory mobile render learning or these more standard tests of their memory function, I think we have found that for MCI versus controls, both tests could distinguish between them, but the effect size that we found for VR, for example, some of these performance measures was much larger. 

So I think that was one of the interesting findings that we've had for these tests. And these need to be replicated in larger samples. But again, like so, what we found was that some of these tests will not show differences between control of some of the standard like cognitive tests, but those who did for some of the VR measures, we found actually better effect sizes. 

Jessica Hagen

So do you think that if you did like cognitive training and virtual reality, you would actually have better outcomes than outside of virtual reality?

Dr. Hadi Hosseini 

That's an interesting question, and I was interested in that for sure. I think one of the challenges that I'm seeing, and I'm doing cognitive interventions in children as well, I think compliance and feasibility is a big challenge, right? Even with a simple computerized test where I would send instructions older adults needed to go on this web page and just go through these games, right? There's so much technical difficulties, and it's just a combination of, you know, different types of operating systems that they would have. Yeah, it's very simple technical issues, still just the technical ability of some of the older adults, right? So from that spectrum, there was a lot of challenge with respect to running that, I think, if it was a VR session, right? 

So I think there have been, like, some tests already developed, for example, for maybe testing allegation or other kind of domains, but in terms of the, I think, kind of training, where you would need to do it more often, if we can somehow circumvent basically those technical challenges, I think we can design tasks to improve a lot of these skills. Yeah, and then that immersive environment would help.

Jessica Hagen

Yeah, it does seem like that would be a pretty big limitation, just like the inability of some older adults to be technologically savvy.

Dr. Kim Bullock 

And I think this is kind of a problem in the field, just the onboarding. And there's so many different devices and so many different systems, you know, this is where I think industry might be able to help us. You know? I know there's this whole serious video game-based instruments who are trying to look at big data of people that are using their products. I'm also hearing too that onboarding and the technology is still a limitation, especially in basic neuroscience. Do you think there's going to be a place for industry and gaming and commercial people to get involved in some of these questions?

Dr. Hadi Hosseini 

I think. So, yeah, I think with advances, specifically in like, you know AR, not necessarily VR, right? So maybe, you know, more augmented reality or mixed reality, and with those advances, I think would be making these technologies plug and play, right? I think with those, I think there's a lot, and I know there's a lot going on in the field already, right? So a lot of larger sample studies or researchers are now doing at home VR tests for different populations, so I'm definitely hopeful, and I'm sure we'll see a lot more.

Jessica Hagen

Yeah, that's true, right? Like AR glasses are much easier to put on than a giant VR headset, yeah. So yeah, yeah.

Dr. Hadi Hosseini

I think the field is growing. There's a lot of, I think, interest, but I think one of the challenges they see is that with assessments, specifically, because we talked about using VR for assessment, is that so you're looking at these kind of, you know, cross-sectional measures, right? So we're just measuring behavior once in their lifetime, and then based on that, we are saying, okay, this person has MCI, or this person doesn't have MCI. I think that's the big challenge. 

So I feel like we need to move forward with having more continuous monitoring or assessment of cognition, even with the VR like we have, I'm looking at my sample and looking at finding like this older adult person who is doing 100% on very difficult navigation tasks. And I'm wondering, okay, what's going on with this person? 

At the same time, there are some healthy older adults that are doing poorly on navigation, and it could be that, from the beginning, they have lower baseline navigation. So I think that's a big challenge that I think, with the help of, like, maybe more accessible VR systems, being able to track this trajectory, I think would be much more informative in terms of picking up or, like, diagnosing diseases. Yeah. 

Dr. Kim Bullock 

Personalizing.

Dr. Hadi Hosseini

Exactly–precision, or personalized psychiatry or neurology.

Dr. Kim Bullock

So would that require wearable VR?

Dr. Hadi Hosseini

You know, I think it would be, like, just more frequent, like, you know, more frequent. Yeah, right. So you have a game for all the adults that they can play, let's say, once a month, even, right. So, and then, yeah, just looking at the trajectory you may be able to pick up, yeah.

Jessica Hagen

Yeah. I wonder how that would work with the extreme progression of technology. If you wanted to do a long term study, we were like, I'm going to study a good amount of people for three years or five years. But then technology progresses so quickly, eventually there's going to be better technology to be able to study. It just seems like they would be, you know, kind of a clash, as far as, like, a long-term study goes with using, you know, this type of technology.

Dr. Hadi Hosseini

Yeah, that's been always a challenge. And for us, like we started with this scanner five years ago, we tried to stick to the same scanner and you know how to look at brain, for example, and then, like, yeah, five years later, you're still behind, yeah. And same for technology, yeah, there's a lot of efforts in terms of how you can harmonize it. Maybe you can switch devices and then still try to harmonize data across different technologies or methods. So it's still possible. 

The field is also like this whole multi site, large scale studies, are, you know, moving along that line, right? But also, I think I would celebrate having more accessible devices, and then there will be all these new studies specifically using those devices. We are now moving toward trying to see how we can assess behavior at home, even like doing brain imaging at home, for example, we have, you know, been working on this headband that we can give to individuals monitor brain function on a daily basis, for 30 minutes a day, or, you know, for several weeks, and then see if you can find any pattern that would tell us anything about, for example, these changes in brain function so on.

Jessica Hagen

That's something that you're currently doing, yeah?

Dr. Hadi Hosseini

I don’t see it as a clash. I would see it as, I think, an opportunity. And as the technology advances, things are becoming smaller and more affordable, like, scalable, right? I think that's going to kind of, like, help with this kind of, like, you know, going toward this precision medicine. 

Jessica Hagen

Yeah, right, absolutely.

Dr. Kim Bullock

Thank you so much. Dr. Hadi Hosseini , your dedication to quality and your forward thinking and optimism about the field is really inspiring. So thank you so much for your time and speaking with us today like we could probably have you again if you're ever interested in and talk about many other things too, but thank you for this initial talk about your work. 

Dr. Hadi Hosseini

Thank you so much. Yeah, it was great. Yeah, being here and it was a great conversation. I hope it benefits the community.

Jessica Hagen 

Yeah, it was really nice talking to you.

Dr. Kim Bullock 

Yeah, I'm sure it will. 

So that's it for this episode of Psychiatry XR. We hope you gained a new perspective on the use of extended reality in healthcare, and thanks for listening. This episode was brought to you by Psychiatry XR, the psychiatry podcast about immersive technology and mental health. For more information about Psychiatry XR, visit our website at psychiatryxr.com, and be sure to subscribe to the podcast and tune in again next month to hear from another guest about XR in psychiatric care. 

You can join us monthly on Apple podcast, Twitter, Spotify, or wherever you get your podcasts.  Psychiatry XR was produced by Dr. Kim Bullock and Jessica Hagen, and please note the podcast is distinct from my clinical teaching and research roles at Stanford University. The information provided is not medical advice and should not be considered or taken as replacement for medical advice. The episode was edited by David Bell and music and audio produced by Austin Hagen. See you next time you.