#44 - AI For Dementia Care

What if artificial intelligence could help make dementia care feel less like a 36-hour day?

Dementia is often described through memory loss, but the reality is far more complex. For caregivers, the hardest part may be the constant vigilance: tracking medications, preventing falls, managing wandering, responding to changing behaviors, and trying to preserve dignity and connection along the way. We explore how AI could support dementia care in practical, meaningful ways, while also asking where the technology could cause harm if it is designed without empathy, usability, and real-world caregiving constraints in mind.

We break down what dementia is—and what it isn’t—across Alzheimer’s disease, vascular dementia, Lewy body dementia, and frontotemporal dementia. Because symptoms and progression vary so widely, assistive technology has to adapt over time, often becoming simpler as a person’s needs change. From there, we look at early detection tools that use machine learning to analyze speech, facial expressions, gait, typing patterns, and everyday behaviors to identify risk earlier and guide screening.

The conversation also moves into daily life: smart pill dispensers, reminders for meals and hygiene, home monitoring, wearables, fall prediction, and wandering alerts. We also examine cognitive support tools like reminiscence therapy, where personalized photos, music, and life stories can help strengthen mood, memory, and connection through conversational AI and voice-based interfaces.

But the promise of AI comes with difficult questions. How do we avoid overwhelming caregivers with constant alerts? When does safety monitoring become surveillance? And what happens when social chatbots reduce loneliness while creating one-sided emotional bonds?

For anyone interested in dementia support, caregiver burnout, digital health, and the future of eldercare, this episode offers a practical map of where AI is already showing promise—and why thoughtful, human-centered design matters just as much as the technology itself.

References:

Assistive Intelligence: A Framework for AI-Powered Technologies Across the Dementia Continuum
Mohapatra et al.
Journal of Ageing and Longevity (2026)

Introduction to Large Language Models (LLMs) for dementia care and research
Treder et al.
Frontiers in Dementia (2024)

Demo: Can Visual Stimulation Enhance Reminiscence-Therapy Chatbot?
Kononovych et al.
NeurIPS Workshop GenAI for Health (2025)

Exploring the Design of Generative AI in Supporting Music-based Reminiscence for Older Adults
Jin et al.
CHI Conference on Human Factors in Computing Systems (2024)

Credits:

Theme music: Nowhere Land, Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0
https://creativecommons.org/licenses/by/4.0/

Transcript

The 36 Hour Day Problem

SPEAKER_01 0:00

The 36-hour day. That's how many caregivers describe the exhaustion of looking after a loved one with dementia. But what if an AI could take the night shift? From sensors that predict falls before they happen to robotic companions that provide 24-7 emotional support. We're looking at how AI could finally give back and what assistive technologies could be on the horizon.

SPEAKER_00 0:35

Hello and welcome back to Code and Cure, the podcast where we discuss decoding health in the age of AI. My name is Vasant Sarathi. I'm an AI researcher and cognitive scientist.

SPEAKER_01 0:46

And I'm here with Laura Hagopian. I'm an emergency medicine physician. And today we're we're we're doing something a little different today, actually.

SPEAKER_00 0:54

Yeah, yeah, yeah, yeah.

SPEAKER_01 0:55

Um first, the first different thing is that we actually have a uh we had a listener ask for this topic.

SPEAKER_00 1:03

Oh, that's awesome. That's awesome. Yeah.

SPEAKER_01 1:05

Keep those requests coming because that's really fun.

SPEAKER_00 1:08

Yeah, yeah. And it also gives us a uh, you know, sort of insight into what, you know, that a what's interesting, but also where are the open questions?

Stages Of Dementia And Personalization

SPEAKER_01 1:17

Yeah. And number two, um, instead of like doing a deep dive into something sort of narrow, we're we're taking this like broad overview today and thinking through, hey, how could AI help here? And I think um if you've if you've never listened to an episode before, that's totally fine. But if you've listened to episodes before, you'll probably see like pieces of other episodes kind of coming in here because there's a lot of opportunities. There's obviously as always, like there are some concerns, but um we're gonna kind of look at the spectrum of how assistive intelligence could work for dementia patients. And this is like a huge problem. I mean, we're talking about dementia affecting over 55 million people worldwide right now, and that number, the prevalence, is expected to triple by 2050 because we have this aging population. So this is like a really, really important topic because it's like, well, how what how are we gonna how are we gonna handle it?

SPEAKER_00 2:20

Yeah, I mean, it's clearly a problem for lots of people and and their families and loved ones. And so, you know, it seems like a place where different types of technologies can come in and be useful in different ways.

SPEAKER_01 2:32

Exactly. And I love that you brought up the fact that of families, because it's it's not just like the the patient themselves that needs help. It's like the caregivers need help, the people surrounding them need help. Right. Um, you know, some people with dementia could be very early on, you know, haven't even recognized they have dementia yet. Some people could need around the clock care. And so there's a lot of um impact that AI could potentially have here. And some of the things have been developed already, and some of them are sort of just more ideas. But I think it's cool to think through, hey, what does the landscape look like? Like what could be developed out for sure.

SPEAKER_00 3:15

Yeah, yeah. I mean, that's that's what this pay. So we have a few papers that we'll uh put in the show notes, but uh uh there's a there's a survey paper that kind of covers the broad spectrum of things because it's you're right, it's the caregivers, the people, but even for the people, there's different aspects, uh dimensions of their um sort of their the cognition, their physical health, and so on, that all can be influenced by dementia. And you know, it's also a disease that progresses, right?

SPEAKER_01 3:42

Right. And so you could think of it in terms of different stages. Like in the beginning, maybe it's really subtle. Like maybe people don't even necessarily notice, although maybe an AI could notice, right? And then over time they start to have more memory impairment, they need more support. Um, and then eventually like routine tasks become more difficult, right? Um, and so as they, you know, have more trouble talking or they have more dependence on their caregivers, they're gonna need more help. And then kind of ultimately in the most severe form, you know, they may not be able to communicate. They may have basically full dependence on their caregivers to um to survive. And so thinking through across the spectrum, hey, how do we personalize this? How do we adapt to these different levels over time is like that's like a very difficult problem. And of course, you couple that with, okay, we're talking about technology here, and now over time, these people may have more difficulty using the technology.

SPEAKER_00 4:46

Yeah, too. I mean, it's not that there isn't any technology out there uh for this. And there have been lots of different tools that have been developed, but a lot of them have been focused on later stages, um, you know, rather than proactive early stage interventions, or a lot of them uh don't account for what happens and the dynamics of the sort of the progressive trajectory uh of the disease itself. And a lot of these different tools do not account for individual differences that we just talked about here. Um, so they have a sort of static notion to them. And top it all off, there is a sort of a fragmented ecosystem, right? You have all these different tools performing all these different functions. And so if some in some ways it might be increasing the burden for people to be able to use it. And then there's the core technology usability aspect, which is do people, are we expecting people to be able to utilize the latest technologies effectively?

SPEAKER_01 5:38

Right? Like, okay, just for an example, like if someone has trouble using a fork, do you think they're gonna be able to use this device, right? Right or whatever. Right.

SPEAKER_00 5:46

So the modality has to be different, right? Absolutely.

SPEAKER_01 5:49

And there's stuff out there. I know there's stuff being studied, and we'll we'll dive into that a bit too, because there are things where, you know, for example, I might have to wear a wearable for it to work, right? But then there's there's other ideas out there, and they're doing research on this already of, oh, can we build this into the home? And then the home monitors them and they don't have to wear that wearable device. Yeah. Right.

SPEAKER_00 6:12

Yeah. Maybe a place to start is to just talk about dementia and what kind of what kind of happens um in the in the patient. I know it's progressive. I know that it goes through stages. Um, and I know that I mean, we think of the dementia as being a brain thing. So obviously you have cognitive impairment, but it's not just that, right? There's more to it.

SPEAKER_01 6:34

Well, that, yeah, you're right. And, you know, we people kind of use the term Alzheimer's and dementia synonymously, but that's not actually the case. Dementia is sort of like that umbrella term um for when people have a decline. And the decline, like you said, is is in brain function. It's in things like memory, it's in um things like language. They have trouble problem solving, they have trouble thinking. Over time, they're going to have um more and more difficulty doing things that you sort of take for granted, like going to the grocery store, making a meal. Um, eventually they may have trouble using the bathroom, those kinds of things. Um, but this umbrella term, it could be from Alzheimer's. Dementia could be from vascular causes. There are lots of different things that can cause dementia. Um, there's there's something called Lewy body dementia. There's another thing called frontotemporal dementia. They show up in different ways, they have different symptoms, different findings in the brain. Um, and and it does become more common with age, but it's not necessarily what we would say is a normal part of getting older. Um, it's just something that we do see see more commonly. So oftentimes it's like, hey, I'm forgetting things. I'm having trouble coming up with the next word, or I'm getting lost, or I'm having trouble planning. Sometimes, especially with something like frontotemporal dementia, you might see personality changes as well. And so there's there's a gamut here. And this is why having sort of individualized or personalized um assistance is so important. Because people may have different symptoms.

SPEAKER_00 8:14

Right, right. These different things might manifest themselves differently.

SPEAKER_01 8:17

Exactly. And they may be at different stages. They might be at the preclinical stage where we want to try to detect it early and see what we can do to intervene, or they may be much farther along, and then you want whatever assistive device you have to be able to adapt to that.

SPEAKER_00 8:34

Right.

SPEAKER_01 8:35

And it's interesting because when I think about adapting adaptive technology, oftentimes I think about things leveling up. Like, for example, the kid are like if we have a kid in school, then you know, they're doing well at math and they do really well at it at edition. And now instead of adding, you know, tens and twenties, they're gonna add hundreds and and four hundreds, and it just progressively gets gets harder for them. Yes. This is kind of the opposite of that, where it needs to level down, level down almost, like make it easier in a way, or and recognize that this level down needs to happen.

SPEAKER_00 9:13

Right, right, right, right. Recognize that's a big piece there.

AI Screening From Speech And Movement

SPEAKER_01 9:16

Yeah. So I think I mean, we've kind of dug into this a little bit already. Maybe we the best place to start, um, and this paper kind of goes through four different dimensions is this concept of cognition, of um, you know, how can how can AI help with either detecting or um stimulating cognition in patients with dementia? Right. And this early detection thing is very interesting because there are subtle things that maybe I would just shrug off if I saw it every once in a while that an AI system could detect. And um, we had an episode about you know, using smiles to potentially detect Parkinson's, right? Exactly. And so the whole idea here is hey, like, what what are people saying? What are their facial expressions? Um, how are they moving around? And so that sort of analysis, how how quickly are they talking, how slowly are they talking, what's the expression, et cetera. That sort of automated analysis can actually, and there's studies that do show that actually works. They can predict who is going to develop dementia using this information and also using information about how someone like uses their technology, like how quickly are they typing, that kind of stuff. Um, how are they walking? And so there's an opportunity to sort of use that technology as like a screening tool to figure out, hey, who's at risk here? Who do we need to keep an eye out for? Is there anything we can do to intervene? Beyond that, though, um, if someone does have dementia and is having cognitive difficulties, uh, I don't know if you've interacted with people with dementia at all, but like sometimes they might go up to their spouse and be like, Who are you?

unknown 11:04

Right.

SPEAKER_01 11:05

Or they may not remember their children, right? And you may tell them, and then five minutes later they ask the same question again and again and again. And so it's uh it's really difficult for caregivers, but there's this cognitive stimulation that they that they need. And if they say, Who are you? they need to know that you are Vasanth. But if it's a different, if it's a different patient with dementia, the answer is gonna be different, right? Right. And someone's um memories, someone's photos, someone's travel history, someone's uh you know, favorite band, those are all gonna be different from person to person.

SPEAKER_00 11:46

Absolutely, yeah.

Reminiscence Therapy With Conversational AI

SPEAKER_01 11:47

But having those memories in a place where you could stimulate someone's cognition is a really, really interesting way to go about this because you really are personalizing and ideally adjusting over time based on what someone's able to do, whether that's like cognitive games, uh, memories, etc. And there's a really cool thing that that has been studied a little bit, which is called reminiscence therapy.

SPEAKER_00 12:21

That's interesting. So in this setting, you have potentially a way to kind of jog the memories. Is that a way to think about this off the patient by presenting them with um their own past and things they liked and disliked or whatever? And it and and that and and research has shown what you're saying is research has shown that that's found to be helpful.

SPEAKER_01 12:44

Um I mean, in theory, you don't need AI for reminiscence therapy. Here's a photograph that you were that when you traveled across the world, or here's some music that you like to listen to, or you know, here's a a video that you enjoyed. Now you take that information that might jog someone's memory and you use digital technology to sort of like augment it or help them reminisce, right? So it's one thing to show them a single photo. What if you like put them, put virtual reality goggles on them and they got to sort of re-experience what it was like to travel to Morocco? I don't know. Right. Or what if you had a conversational, like a chat bot that could adapt and have conversations with someone based on that? And what if it did adapt over time? So it's monitoring, say, your facial expression, seeing what that looks like. Is it joyful? Is it scared? Is it sad? So that it can adapt what it's showing people. Now, this is like, you know, this is like ideas of how it could work, but there is some evidence that that this stuff does work in in small subsets of patients where they've gone and said, hey, let's try to put this conversational agent where let's program something into it. And I uh it and it's definitely been shown to help with people's mood, to help with their cognition, et cetera, to be able to reminisce and relive those memories from the past.

SPEAKER_00 14:16

Yeah. And you know, the the the space of it's a very sort of novel or sort of space where there isn't that that many, there aren't that many tools out there. There's very few, very little research done on using large language models and chatbots for the purposes of uh reminiscence therapy and generally for dementia. What's interesting about it is also, you know, we've talked about mental health chatbots, right? Chatbots that help and we've talked about the chat, we've talked about the challenges in those settings where people are reaching out to them and having all these issues where a therapist might deal with that patient differently from how a chatbot should deal with that patient. Right. Yeah. And in this case, it's it the use the user spaces is slightly different. And you you might have some of those issues might not be the case, might not be there anymore. But at the same time, you do expect a certain um if you are teaching someone on how to to to talk to a dementia patient, there are certain rules, right? There are certain um I don't even want to say rules, but certain ways of talking and certain uh you know things you can say that can make them feel better. I mean, in addition to presenting them with the right photos and the right music and the bands and all that.

SPEAKER_01 15:27

Yeah, like being, I mean, uh with any type of patient, you want to make sure that there's like empathy infused into it, of course. And uh you can imagine it's easy to get frustrated when you're asked for the 17th time, who are you today, right?

SPEAKER_00 15:40

Yeah, yeah, yeah. And you know, in one of the papers we have, they provide an example of a prompt for an agent that um for a chatbot that can do some of this. And um, you know, they start off by saying, You are um a reminiscence therapist facilitating therapy sessions through conversation. Um, and then they said, um, you know, always greet the user warmly to set a positive uh tone and ask the ask them their name, um, refer them to an emotional wheel, and this is sort of another tool that can help them kind of figure out the right feelings and where they are right now. It's sort of a mood assessment of some sort. But they also, you know, in the prompt, you also want to make sure that there are some certain safety checks in place. Um, you know, uh if the user consents, ask warmly what was the happiest time in their lives, uh, offering options like childhood, adolescence, school years, a university period, you know, and focus, and this is part of the prompt still, right? Uh ask about a most prominent memory from this period, the user's feeling in that moment. So in this particular instance, um, they're sort of collect using a collecting memory. Collecting memory. But of course, if the patient has has um you know sort of gone past that point where they're remembering things or not remembering things, uh you could easily imagine such a tool being used with a caregiver. And in in in my head, the way I think about this is that the caregivers are doing a lot of different things. And if a tool like this can sort of unburden them from having to um do this, you know, maybe just do it once, right? Get me, get me some set of memories, give me a collection of photos, give me a collection of things, and you do it once, and now you have a better understanding of the patient, uh, or the AI system has a better understanding of the pre patient and can use that as a set of collected memories to draw upon later.

SPEAKER_01 17:24

Yeah, and uh there are some definite benefits to that, right? Because, well, the chatbot doesn't get frustrated when it gets asked the same question 17 times in a row, right? Well, yeah.

SPEAKER_00 17:33

The thing we said was it's maybe a useful thing for having a certain amount of therapist friction, we said would be useful in certain mental health conditions. But in this case, you actually want the chatbot to be just always positive, right?

SPEAKER_01 17:46

Um and and on sort of 24-7, right? You want them to be able to sort of access it. I there's there are probably there are, I'm sure, exceptions to that, but like in general, it makes sense that in order to unburden a caregiver, you have something like this available. Right. And it could not only help with you know agitation and and mood and things like that, but it could also potentially detect it. And so mental health, both emotion detection, and then we've already started to talk about how there could be social support here, right? That patients with dementia often experience loneliness, and that's not surprising, right? Um, and that could vary over time, but that's a major contribution um where it could unburden caregivers also.

Loneliness Support And Ethical Tradeoffs

SPEAKER_00 18:40

Yeah. And I mean, there's obviously challenges. So, you know, this reminds me of uh one sort of body of ethical issues, which is yes, we want loneliness support, and maybe um, you know, patients begin seeing the chatbot as a friend that they can confide in and share their emotions. But you know, I think it's worthwhile sort of keeping in mind that this emotional bond is unidirectional. It is not the case that the machine is developing an emotional connection to the patient, like a human would, right? Right. Like a human caregiver would. There is a distinction there. And whether the degree to which that matters or not is is a different question. But it is something. And the the you know, that there's actually a great movie called Robot and Frank from a while back about a patient, an elderly uh man who had a robot helper in the house, and he really developed this deep bond with this robot that was just doing all the household chores to the point where he didn't want the robot to be doing some of the chores. And he called in his daughter to do you know wash dishes and things like that, and she said, I can just turn the robot on. And and he was not he was not happy with that. And he developed this deep, deep connection with that. And there's a lot of interesting um stories around this. I mean, there was a recent episode of Last Week Tonight with John Oliver on AI chatbots and about how people develop these deep, deep bonds, and it's unidirectional, right? Because you're bonding with the system. The system isn't bonding with you, right? And exactly.

SPEAKER_01 20:05

That's how it can alleviate loneliness, it can provide social support, it can be there for conversation, for engagement. Yes. It's available all the time. Yes, but it's not the same as having a human there.

SPEAKER_00 20:17

Yes. And I I don't know to what degree that's been studied with chatbots, but that thing's something that should be kept in mind as well. Um, but it does seem to have seem to have a lot of promise in that regard. I mean, you could even have these chatbots call in, right? There's no reason that it has to be a text-based interface. Um, you know, you could easily have it just call the patient every, every every day at certain times and talk to the patient, and and that's the interface, in which case you take out a lot of the issues with um, you know, typing and using a device and such.

Independence Through Smart Homes And Sensors

SPEAKER_01 20:45

For sure. I think that makes a lot of sense. And there's even um there's even something to it about, hey, maybe we're just recognizing what the expression is on someone's face if they're not verbal anymore at this point, too. So there's definitely some opportunities there. Um, another place which I think is big is sort of this like uh staying independent, doing what you need to for your physical health. Yeah. Um, and that is around doing basic things, activities of daily living, um, going to the bathroom, showering, those kinds of things that you just kind of like need to do need to do to survive, eating, like whatever. And then there's um instrumental activities of daily living, which require uh a little bit more. They're a little bit more difficult to do, things that require planning, like prepping a meal or um figuring out how you're gonna go from point A to point B or shopping. Those are things that are a little bit harder to do. And so there are lots of opportunities to automate those things, like having um, you know, a pill dispenser that's smart that says, Hey, Vicanth, it's time for you to take your 4 p.m. Pills and it dispenses them for you and it sort of monitors whether or not you've taken them. Those that actually exists already. But um I think there's a lot of automation of sort of routine things that could happen that could let people be more independent, whether that's like controlling the temperature in the house or turning the lights on and off at certain times of day, um, prompting someone to brush their teeth or to wash their hands after they've used a bathroom or before they're going to start a meal. These are all things that could be integrated in here. And wearable devices is another example where it's like, oh, do we think this person might fall? How are they walking today? Or if their wearable device isn't documenting any movement for a while and they're usually moving at this time, is there something wrong? Um, or did they wander? Did they go outside of, you know, maybe they're allowed to walk in their neighborhood? Are they farther outside their neighborhood than they should be? And so this concept of like, it doesn't necessarily have to be wearable devices, right? It could be like a home-based device. But the idea would be hey, can we can we sense this person? Can we understand what they're doing, where they're at, how they've slept, um, you know, what's what's the temperature in their home? Have they exited through a door? All of those things could make monitoring better. You could also imagine that it could make it worse. What if you're getting like a million alarms a day? Door open, door closed, like, oh, they let the, you know, they want to go, we get a package. And so that's very different than them leaving the house and wandering. Yeah. And so you could, you could totally see alarm fatigue um setting in for something like that. And also you could see AI potentially helping with that, trying to understand, okay, why did the door open?

SPEAKER_00 23:48

Yeah. And there's also, you know, in addition to alarm fatigue for the patient themselves, there is a they, you know, there is a certain value in promoting a degree of autonomy and and maintaining a degree of dignity for the patient. Like you don't want, you want to strike that right balance, right? You, you know, yes, you can do things for the patient, but having them feel like they are in in somewhat of control for certain things, that can have value as well, especially for dementia patients. And so that's a hard trade-off to strike as well for an AI system to exactly understand where that boundary is. Like, do I do this or do I just like give them a couple of hints and have them do it and feel good about having accomplished something?

SPEAKER_01 24:24

Yeah, totally. And I think this autonomy thing is kind of it's like very difficult in this context to figure out, you know, where's the line? Because the line's changing.

SPEAKER_00 24:33

Yeah.

SPEAKER_01 24:34

How autonomous can be someone can be is going to be changing over time. But this concept of like, hey, you know, what are also what are we doing with all this data? Is it private? Is it how is it stored? Um, all of all of these issues that are kind of widespread with with the use of AI, with the use of LLMs throughout these technologies, are all going to be present here too.

SPEAKER_00 24:59

Right, right, right.

SPEAKER_01 25:01

The the last area in this paper is around caregiver support. And I think this is like some of the things that we don't talk enough about, right? It's, you know, if someone has dementia and you're taking care of them, first of all, there might be multiple people taking care of them. You you may have some hired help, right? Or some help provided by your insurance, but it that's not around the clock. And so these it's very common for people to who are caregiving to get overwhelmed, right? There's like so much going on. There's so much to do. It's a 24-7 job, in addition to whatever else you might be doing in your life. Yeah, you may not be able to um take care of your own health. You may withdraw socially, you might be exhausted. There's there's just so much going on. And so it's like, well, if there are ways for us to support caregivers with this, that's amazing. And we talked about like alerts already, you know, oh, someone has walked outside the radius they're supposed to go. Um, there's this constant vigilance that people have to have. And it's like, well, can we relieve that if we have them, you know, using sensors with wearable data? Or and can we also have an have a personalized AI chat bot for the caregiver too? Right. If the caregiver has a question or they want to know what to do in a certain scenario, and they can't reach the doctor's office, or they can't reach the doctor's office quickly, maybe there's an opportunity there because these systems can be trained on tons of data. There are so many dementia patients, and then give, you know, learn what the most appropriate answer would be and personalize that for the caregiver and for where the patient is in their dementia trajectory, too.

SPEAKER_00 26:52

Yeah. No, that's that totally makes sense. And, you know, having that, this is where I think the issue of all the tools that are out there being so fragmented is especially a bigger burden. So the more tools you add to this, the more capabilities you add, in some weird way you're actually adding more burden for the caregiver to now, there's another signal to track.

SPEAKER_01 27:09

Yeah, we can't have that. We can't have that in the world. We cannot have that. We would have to all be integrated and together. And this is why, you know, this is it's really cool to review this, but a lot of the stuff that we're talking about today is like in prototype mode. It's, you know, um maybe been piloted. A lot of it is not available commercially right now. And and if it is developed, it's like, yeah, it has to be done in a way that it's integrated with all the other stuff. Because what you don't need is another thing to do when you're a caregiver. You need fewer things to do, you need the burden taken off of you. Yes. And so, you know, something like a smart pill dispenser, that's already, that's already available. But having, you know, AI speech-based screening, that that's something that's that's not wide, that's not really used in a widespread way right now. Right. I think there's definitely opportunity here, but it it has to be thoughtfully deployed. And the personalization is really important, and the dynamic nature of it would be really important too, because what a patient with dementia looks like today and what they look like three months from now could be totally different. And the system needs to be able to identify and adapt to that.

SPEAKER_00 28:23

Yes, absolutely.

What Needs To Happen Next

SPEAKER_01 28:25

All right. Well, that was that was a lot about dementia. I think there's a lot of opportunity for AI to step in here and assist, whether it's with cognition, with mental health, with physical health, uh, or or with caregiver support. A lot of this is theoretical right now or in that sort of pilot phase. But I think this is a really important problem because the dementia population of patients with dementia is large and it's growing. And the burden on caregivers is large and it's growing. And so we need, we have this opportunity to develop solutions that could help on all these fronts. And I think that's that's something that we need to dive into more.

SPEAKER_00 29:05

I agree. Thank you for joining us.

SPEAKER_01 29:07

We'll see you next time on Code and Cure.