- [Announcer] "Your Fantastic Mind," brought to you in part by Sarah and Jim Kennedy.

(upbeat electronic music) (upbeat electronic music continues) - Welcome to "Your Fantastic Mind."

I'm Jaye Watson.

Strides in Alzheimer's research continue at an astonishing pace.

We have two drugs approved to treat people who are at early stages of the disease: blood tests to diagnose, but not yet approved by the FDA.

And we have agency; lifestyle interventions, things we can do that cut the risk of cognitive decline almost in half.

This week, we take you into the classroom to show you how aspiring medical students are trained to understand cognitive decline and the people impacted by it.

We look at the connection between pollution and Alzheimer's disease and we show the promise of artificial intelligence to overhaul the system so people get diagnosed earlier, diagnosed more easily, treated quickly, and even switch to a different treatment because AI can tell if it's not working.

We begin tonight with a program for people diagnosed with cognitive decline.

A living laboratory that is building community, teaching skills and habits and helping its members hold steady.

(light electronic music) - One unit of olive oil here.

I would shoot for a starting off with a tablespoon.

- [Jaye] 62-year-old Danny Van Winkle and his wife Heidi love cooking class.

- Well, there's a first for me.

We never thought we would learn this much in such a short period of time.

We have learned so much.

- [Jaye] It's been four years since Danny was diagnosed with mild cognitive impairment at 58 years old.

- They did a lumbar puncture because that is the only way really to diagnose Alzheimer's.

So that test came back positive, so we were devastated.

- So when we're talking about our awareness... - [Jaye] The Van Winkles were referred to Emory's Cognitive Empowerment Program, a comprehensive lifestyle program for people who have a diagnosis of mild cognitive impairment and their care partners.

Director of the program, Amy Rodriguez.

- So we are in the business of delivering a structured intervention that helps people make positive lifestyle changes in five domains.

- [Jaye] Those five areas are cognition, physical activity, nutrition, wellbeing, and functional independence.

- So driving is by far the most complicated activity of daily living that we do.

- [Jaye] One full day a week, participants work in the five areas in community with others with the same diagnosis.

(Danny and Heidi laughing) - The really big thing for me is the other people we're with.

(co-participant laughs) God, we made some great friends and we all feel like we've known each other for forever.

(participants laughing) You're gonna be a movie star.

- That's right.

- If you let yourself.

- If I don't remember something, it's on the board.

- [Jaye] 75-year-old Judy Hood and her partner, Barbara French, are program graduates.

- We play games.

- Oh, that's right.

- The game's what we play, every day.

- Yes, we do.

- Every day.

- Yes, we do.

Keeps everything going.

I think you start first.

(dice clattering) - [Jaye] They're home and lifestyle routine came from strategies learned in the program where they stayed in touch with fellow members.

- Just not alone.

It's a lifeline.

For me, it's just a lifeline.

It teaches us to live a moment at a time.

Right Judy?

- Yes.

The group that we, that I was in, it was a very short time before we were all a family.

And everybody's hugging everybody when they come in, and you know, you just love these people.

(gentle music) - [Jaye] Launched in January of 2020, the Cognitive Empowerment Program is successfully teaching people with memory loss, new strategies and habits.

- [Amy] We do a lot of repetition within the program.

(Judy laughing) - [Jaye] Care partners reinforce at home.

They can also access everything they've learned in an app created for program members.

- [Participants] One... - [Jaye] Known as a living laboratory, the research-based program uses measures of cognition, everyday functioning and wellbeing at the beginning and end of the program to track results.

These measures are important for building the evidence for lifestyle interventions in mild cognitive impairment by demonstrating the impact on disease progression and on factors that influence quality of life.

- What we've found so far in looking at our outcomes is that after a year of participation in the program, people are stable.

You know, their cognition remains stable, they're stable in their everyday functioning.

- [Jaye] Members also report increased empowerment and self-sufficiency and decreased depression and stress.

Decreasing depression is important because it is one of 14 modifiable risk factors for dementia.

In 2024, the Lancet Commission updated its report on dementia prevention, identifying those 14 risk factors that, if addressed, could potentially prevent or delay up to 45% of dementia cases.

These risk factors span the stages of life and include low education, hypertension, hearing loss, obesity, excessive alcohol consumption, traumatic brain injury, high LDL cholesterol, smoking, depression, physical inactivity, social isolation, diabetes, air pollution, vision loss.

The body of research into lifestyle interventions to prevent cognitive impairment and dementia is growing.

And the FINGER study, Finnish Geriatric Intervention Study out of Finland, was groundbreaking.

Conducted on over 1200 healthy adults at risk of dementia between 60 and 77 years old.

They were randomly assigned to either a control group where they got advice about a healthy lifestyle or to an intervention group where they did supervised aerobic exercise and strength training multiple times a week, followed a Mediterranean-style diet, reduced sugar, processed foods and unhealthy fats and received nutritional counseling, and monitored and managed their blood pressure, cholesterol and blood sugar.

After two years, the intervention group had 25% better cognitive function than the control group.

The intervention group's memory function improved by 40% and processing speed by 150%.

While lifestyle interventions are powerful, there are only a handful of programs like Cognitive Empowerment across the country, but we can adopt these positive lifestyle changes to reduce our own risk.

It's never too late.

- What I thought was like so brilliant, when they started the whole like smoking cessation thing, is I would say, in just one year after you quit smoking, this'll be the outcome.

In five years, now you're at lower risk for, you know, heart disease, by 10 years... And so it like breaks down.

Like, it's never too late to start.

(participants laughing) - We're very lucky because I think we're gonna have Danny longer than what we first anticipated.

We are leaving with hope.

We're leaving with Danny living longer independently.

(gentle music) (light electronic music) (traffic droning) - [Anke Huels] Research over the last decades has really shown a strong link between air pollution and the risk of developing Alzheimer's disease.

- [Researcher] So this is the wet, the wet tissues.

- [Jaye] Anke Huels is a researcher at the Rollins School of Public Health at Emory University.

- We started a collaboration here with, with a brain bank at Emory.

- [Jaye] For her research into the impacts of pollution on the brain.

She examined 256 brains of people from different areas of Atlanta.

- Based on the last address where they lived before they died, we assigned air pollution concentrations to that address to reflect what kind of air pollution were they exposed in the years prior to their death.

We found that participants that lived closer to major roads before they died, and were therefore exposed to higher levels of air pollution, they had a higher amyloid plaque burden in their brains.

(horn honks) (traffic droning) - [Jaye] Huels says the major source of air pollution in Atlanta comes from traffic and that levels are highest on the south side of the city.

- The intersections between all the major highways.

- [Jaye] There is mounting evidence of air pollution's effects on the brain.

In the early 2000s, shortly after the UN deemed Mexico City the most polluted in the world, researchers performed autopsies on the brains of stray dogs and found pollution particles in their brains.

And that's when researchers began to realize that some of the smallest particles can be the biggest offenders.

The most often mentioned is PM 2.5, particulate matter smaller than 2.5 micrometers in diameter.

It usually comes from vehicle exhaust and smoke or dust.

It is tiny, 30 times smaller than the width of the average human hair.

It can stay airborne for a long time and enter our bodies.

- These small particles, PM 2.5 and even smaller particles like nanoparticles, they again, they enter your nose when you breathe in and then they can actually enter your brain through the olfactory nerve.

PM 2.5 is just defined by its size.

So it's a mixture of many different chemicals.

And many of those chemicals have been shown to be neurotoxic.

So once they've entered your brain, they can cause direct damage to the brain and therefore also cause disease like Alzheimer's disease.

(horn blaring) - [Jaye] Before you panic and move to the middle of nowhere, it's important to remember that the impacts of air pollution are over a lifetime.

Also important, lifestyle factors that prevent cognitive decline also protect against air pollution impacts.

And your status in life is also protective.

Three years ago, using data from an online survey of 20,000 people who were asked if their memory was worse than the year before, Huels and her team found that in high air pollution concentration areas south of Atlanta that are socioeconomically disadvantaged, people there have higher rates of cognitive decline than people in high air pollution areas that are in higher income neighborhoods with strong infrastructure.

- Those people are kind of protected from the adverse effects of air pollution.

So we really didn't see any effects of air pollution on cognitive function among the participants that live in these neighborhoods in the metro Atlanta area.

So more in the north of Atlanta.

(light electronic music) - Right now, the pathway through a healthcare system for somebody with dementia is a journey.

It's a nightmare, just like it is for anybody with any complicated chronic disease.

- [Jaye] He's right.

This is Dr. Allan Levey, head of the Emory Goizueta Brain Health Institute.

- It's really tragic because most people still get delayed diagnosis, no diagnosis.

Rural America, you know, underrepresented minorities, they've such poor access to healthcare that that's number one.

- [Jaye] The plan: to end excruciating wait times and delayed diagnoses and treatments is by using artificial intelligence overseen by humans.

We're going to hear from four leaders in AI and Alzheimer's research who are going to paint a picture of how AI could do away with healthcare bottlenecks and lost time that can make it too late to receive treatment.

Let's start at the beginning.

- So we are in the CODA building.

The CODA building is in Midtown Atlanta.

It's like, I like to say it's the temple of AI.

So we are funded by the National Science Foundation.

We are one of the 27 AI institutes across the country.

- [Jaye] Pascal Van Hentenryck is the director of the National Science Foundation Artificial Intelligence Institute for Advances in Optimization.

- This is one of the labs that we have.

My name is Pascal Van Hentenryck- - [Jaye] And the head of Tech AI, the AI hub at Georgia Tech.

- So this is part of AI for mobility.

- [Jaye] Van Hentenryck and his team have a heavy lift in this massive endeavor in collaboration with Emory, to improve and accelerate access to assessment, diagnosis, and treatment.

- We'll do this much faster, right?

So the time now to get an appointment is around a year, it's gonna be around 14 months, is like, you know, in the next 10, 20 years.

So it's not acceptable.

So if you don't change the system, people are not gonna be diagnosed early enough.

So you need to change the way the system is working and I think the really huge opportunity that we have is that AI can give you that aspect of it.

(intriguing synth music) - [Jaye] Van Hentenryck and his team will use large language models, which is where AI learns billions of pages of information on a certain topic.

After it learns this massive amount of data, it can answer questions, summarize information, and engage in conversations.

- I think we wanna show you the workflow.

- [Jaye] In the case of Van Hentenryck's research, the focus is training AI on decades of patient data.

- What we need to do is take a lot of data, as much data as we can from patients over the last 30 years.

So once we have that body of knowledge, we want to use generative AI, so this is the technology which is in large language model, and we want that technology to generate synthetic patients.

These synthetic patients are gonna give us, you know, we can generate now artificial patients.

Think of virtual patients that can help us envision how to make the system work effectively.

Because now we can simulate the system at very large scale with patients that are really modeling what the disease is gonna do.

- [Jaye] These artificial patients will talk with AI conversational agents, chatbots, in order to help doctors identify which of them could benefit from medication early in the disease progression.

These artificial patients created from real life data will also help Van Hentenryck and his team learn the problem spots, the bottlenecks, the slowdowns that jeopardize care.

Then, partnering AI with physicians, they'll create a new path forward to speed the process along.

AI will be crucial at every step of the way, including the beginning.

- [AI Conversational Agent] Let's begin.

Can you tell me what brings you in today?

- Yeah, I've been more and more concerned about my memory.

- [Jaye] Drew Breithaupt is an Emory cognitive neurologist and clinical researcher who is exploring how artificial intelligence can enhance care delivery.

- Before I was misplacing my keys- - [Jaye] And he's talking to a conversational agent.

- [Conversational Agent] Have you found yourself forgetting events like appointments or meetups with friends or family?

- There have been some times where my son talks about us having coffee.

Usually this is two or three weeks ago and sometimes I really struggle to remember.

- [Conversational Agent] Noticing gaps in memory, especially with events from just a few weeks ago, can be frustrating.

- [Jaye] The conversational agents are becoming advanced and good at their jobs.

- The words that it puts together are very understandable.

It can explain complicated topics, including complicated medical topics in a way that's easy to understand.

It can adjust to the patient's level of education.

It can also speak to a patient in a very inviting and empathetic manner.

And when people are blinded to what ChatGPT is saying to them in medical terms and what a physician is saying to them in medical terms, people prefer ChatGPT.

- [Conversational Agent] Has this happened more than once?

- [Jaye] Conversational agents can interview, gather information and data, create a summary for the doctor, maybe even a suggested diagnosis based on this information, - [Conversational Agent] We'll work together to understand what's going on and find the best way to help you.

- Physicians everywhere are struggling to find the time to ask the questions they need to ask for some of these more complex conditions.

There have been some times where- - [Jaye] There's a lot of research into how AI can assist with diagnosis using the patient's speech, how long it takes them to respond, how they use or forget words or repeat themselves.

All of this can be done before the doctor ever sees the patient.

- I am wowed by some of its abilities that I can see right in front of my eyes.

With being able to make complex diagnoses that took me 10 years to learn training.

It's kind of, that was another gut punch.

- [Jaye] Then there are self-administered AI tests.

- It could be using their smartphone, it could be taking a formal test on a laptop or a iPad.

You know, AI is enabling us to very rapidly develop tools that people don't mind doing that will give great insights into their cognition.

- [AI Conversational Agent] Tap the start button when you are ready to begin - [Jaye] Tests the Linus app provide cognitive testing for patients on iPads in their physician's office, including scoring how well they do.

This is another way to determine who needs more testing or to be referred to a neurologist.

- Yeah, like I'm doing now, I'm looking into this camera and it's looking at where my eyes are.

We look at that.

- [Jaye] Researcher Gari Clifford wears a lot of hats.

- Well, I'm a physicist who became an engineer who became a informaticist working on developing and applying artificial intelligence and signal processing to anything in healthcare.

- [Jaye] Clifford and his team have created an eye tracking app for cognitive decline.

People are shown a series of pictures and a camera tracks where their eyes focus on the screen.

- We provide them different images and we use the webcam on the iPad to track what they're looking at in the image, and then we show them a series of images and then we delete or add something to those images in a very subtle way.

And if you are slightly cognitively impaired, you are less likely to notice those changes.

- [Jaye] The goal of the app is to put it into broader use at clinics.

- [Patient] Nothing really, as I say.

- [Jaye] Clifford is also using AI to help detect depression, a common symptom in those with cognitive decline.

- [Patient] I used to go out with my friends.

- [Jaye] Clifford and his team discovered that changes in the brain when someone is depressed correlate to subtle changes in facial expression and that AI can accurately detect if someone is depressed.

- We know that just gait alone, the way that you walk- - [Jaye] And then there is how people move when they have cognitive decline.

Clifford and his team installed cameras in the ceiling of the Cognitive Empowerment Program space at Emory.

The program is for people in the early stages of cognitive decline, and for their care partners.

- Every single one of these cameras is running digitization and analysis algorithms in real time.

We put them in the ceiling in an unobtrusive way so we can monitor everybody's movements within the space, because what we're looking for is how do people interact with each other?

Do they socialize?

Because we know that social interactions start to diminish when you become cognitively impaired.

And what we're looking for is whether the therapies are leading to increased social activity.

So what we're looking at here is a heat map of just where everybody spends their time.

- [Jaye] The cameras have computers attached.

Using AI, they run real-time stick figure digitization.

- You don't even need somebody having a wearable.

You don't need them to sit on a call.

They don't have to be actually interacting with the clinician themselves.

They can just be walking around this space and just from the way that they walk around the space, we can identify their level of cognitive impairment.

- Let's just focus on what we have to do.

Explainability really is important.

It goes back to what I was saying last time.

I've always been a biomedical engineer, but my interest in biomedical engineering is development and application of artificial intelligence or AI and machine learning algorithms to address problems in healthcare and medicine.

Think of it also in terms of proportionality.

- [Jaye] Anant Madabhushi is the founding director of the Emory Empathetic AI for Health Institute.

- These are some of the results.

- What Madabhushi and his team have discovered could change how and when people are diagnosed with Alzheimer's disease - For progression-free survival.

- [Jaye] And the answer is in our eyes.

- We are gonna be looking at images of the eyes.

What the machine learning algorithms are looking for.

They're looking for subtle cues that reflect the differences and the twistedness, the tortuosity, the angularity, the branching patterns of the vasculature in the eye.

And it learns that there are differences, significant differences in those patients that have Alzheimer's from those who don't.

- [Jaye] Madabhushi and his team used machine learning to train AI on the vasculature of the eye.

- And just by repeatedly showing the machine learning algorithm several examples of the vasculature in Alzheimer's patients and in patients who don't have Alzheimer's, the machine learning algorithm is able to find those specific features that are different between the Alzheimer's and non Alzheimer's patients.

Looking at almost 1400 patients from within the Emory Brain Health Institute, looking at images of the eye, we've been able to demonstrate that we can find with a high fidelity, a high accuracy which patients have Alzheimer's versus those patients who don't.

Being able to connect the oculomic patterns- - [Jaye] How AI can determine who has Alzheimer's or not, has to do with the differences in the appearance of the vasculature in the eye.

- [Dr. Madabhushi] Vessels in the eye in Alzheimer's end up looking more twisted, more tortuous.

These are very subtle patterns.

These are patterns where if you're visually looking at it, you may not be able to truly discern the more chaotic arrangement, the more chaotic architecture.

But that's where the machine learning algorithm basically helps.

(soft synth music) - [Jaye] The goal is to use AI to diagnose people as early as possible so they can begin treatment.

And to also use it to identify early on if treatment is or isn't working.

PET scans and MRIs are expensive and not always available in rural America.

An inexpensive non-invasive eye exam could provide answers.

Madabhushi's team is seeing that if the treatment is working, the vessels of the eye will begin to look normal again, less twisted and tortuous.

If the vessels don't improve, then the treatment isn't working.

These new medications can be a huge financial burden and halting treatment that isn't working will give someone a chance to try another treatment without losing even more time and money.

- I think AI is gonna be an essential partner to the humans to devise new care models.

- [Jaye] The decades-long journey to cure Alzheimer's disease is evolving.

Technology is more than a tool, it's a partner, pursuing a future where memories are no longer lost.

(hopeful ethereal music) (light electronic music) - Let's go ahead and get started.

Pathophysiology of dementia, I think, is the way it's listed in the curriculum.

So these conditions, mild cognitive impairment, mild to moderate, severe dementia- - [Jaye] For half a day during their second year of medical school, students will learn from leading experts and researchers in the field of Alzheimer's disease.

- [James Lah] This is the Montreal Cognitive Assessment- - [Jaye] Neurologist James Lah- - that we'll see in the coming years- - [Jaye] associate director of the Emory Goizueta Brain Health Institute.

- We do a series of lectures going over basic clinical and pathological features - [Jaye] Every year, the favorite part of the day is this.

- This is Jerry and Rachel Cook who are here with us today.

- As I begin to lose more function- - [Jaye] When the students get to hear from and ask questions of a couple living with impacts of cognitive decline.

- What I have noticed is a very slow but very steady decline.

- And how you're balancing that emotional toll- - [Jaye] Rachel and Jerry patiently take questions.

- It's hard to think too far into the future, and so I don't.

- My speech, for example, is finally being affected.

(walkway judders) (padlock clicks) - [Jaye] 80-year-old Jerry Cook is in the early stages of dementia.

(saw whining) At home, Jerry still loves woodworking in his shop.

- [Jerry] Ready to do a crossword?

- He and Rachel still travel- - "Be up in the air."

- [Jaye] But there's no escaping the truth of a disease that moves in one direction, and he has seen it up close.

- Just like my sister, my mom, I don't wanna be like that.

- [Jaye] Jerry lost his mother to Alzheimer's and his sister is at an advanced stage, so he and Rachel focus on the present.

- We're really enjoying right now, you know, the fact that he is where he is right now and we can enjoy as much of that as we have.

- [Jerry] It was just like four of us, four guys to start with.

Well, it didn't take too long before I was the only guy left.

And ever since it's been me and several women.

(tap shoes clack) - Shuffle, step, heel, step.

- [Jaye] A few years ago Jerry took up tap dancing.

He loves to dance and he wasn't going to let his diagnosis deter him.

(upbeat tap music) (shoes tapping) - [Instructor] Five, six, seven, eight.

- [Jaye] At the end of the year, Decatur School of Ballet puts on a performance and Jerry is front and center.

♪ I went to a bar the other night ♪ It's convenient, they put me in sort of the centerpiece for each of these performances and they're working around me and I'm doing something that means I don't have to dance quite as complicated as fast as they do.

(upbeat jazz show music) But it's great, and so I'm sort of like the center of attention of the thing, it's really funny.

I've experienced loss of energy more recently too.

- [Jaye] And that's the goal of this class, to teach students that one day, they won't be just treating a cluster of symptoms, they will be treating a person like Jerry.

- Not somebody that you've seen one time in a skilled nursing home or not the brain of somebody that you saw at a brain cutting or pathology slides that you examined.

They're somebody who used to work in IT at Emory University and does tap dancing and is a very cool guy.

(tap shoes clacking) - Thanks for joining us this week.

See you next time on "Your Fantastic Mind."

(upbeat electronic music) (upbeat electronic music continues) - [Announcer] "Your Fantastic Mind," brought to you in part by Sarah and Jim Kennedy.