Artificial intelligence in healthcare: opportunities and challenges | Navid Toosi Saidy | TEDxQUT

Summary

Artificial Intelligence (AI) is revolutionizing healthcare by personalizing treatment and enhancing hospital efficiency. AI models, trained on vast data sets, can predict patient outcomes and recommend tailored treatments. In cancer care, AI aids in diagnosing and identifying optimal therapies. However, current regulations aren't equipped for AI's dynamic learning capabilities, potentially hindering its full potential. New frameworks are needed to ensure AI's safe, effective integration into healthcare, promoting personalized health advice for all.

Takeaways

• 🤖 **AI as a Lifesaver**: Artificial intelligence has the potential to save lives and enhance healthcare for millions.
• 🏥 **Personalized Healthcare**: AI helps in personalizing care delivery, making hospitals more efficient, and improving healthcare access.
• 🧠 **Learning Models**: AI operates by training computer models on large datasets to make decisions or predict outcomes.
• 👨‍⚕️ **Better Decisions**: AI models can inform doctors about the most effective treatments for individual patients.
• 🧬 **Genetic Insights**: AI can analyze genetic information to aid in diagnosis and treatment, especially in complex cases like cancer.
• 🩺 **Cancer Care Innovation**: AI models are being developed to streamline cancer diagnosis and treatment planning.
• 🌐 **Global Impact**: The application of AI in healthcare is a global endeavor, aiming to improve outcomes for diverse populations.
• 📚 **Regulatory Challenges**: Existing regulatory frameworks are not designed for AI-based medical software and may limit its potential.
• 🔄 **Evolving AI**: AI software differs from traditional software as it learns and evolves, which current regulations do not fully accommodate.
• 🌟 **New Regulatory Frameworks**: Emerging proposals for regulation could unlock AI's full potential in healthcare if implemented correctly.
• 🌈 **Diverse Data Sets**: It's crucial that AI training data represents the entire human population to avoid bias and ensure accuracy.

Q & A

What is the potential of AI in healthcare according to the script?

AI has the potential to save lives, personalize healthcare delivery, make hospitals more efficient, and improve access to healthcare by providing accurate decision-making tools.

How does AI work in the context described in the script?

AI works by educating a computer model using complex and large data sets, allowing it to learn and make decisions or predict outcomes when presented with new data.

What role does AI play in personalized medicine?

AI helps in personalized medicine by analyzing data from thousands of patients to determine the most effective treatment for an individual based on their unique conditions.

How is AI being used in cancer diagnosis as mentioned in the script?

AI is used in cancer diagnosis by consolidating information from blood tests, X-ray images, and genetic information to provide accurate predictions of diagnosis, treatment options, and prognosis.

What is the significance of the AI tool developed in Brisbane for cancer patients?

The AI tool developed in Brisbane can accurately identify the primary site of cancer, empowering doctors to provide specific treatments that are likely to work for the patient.

Why is the existing regulation framework a challenge for AI in healthcare?

Existing regulation frameworks are not designed for AI software, which can learn and evolve over time, unlike traditional software. This can limit AI's potential to improve its functionality or accuracy.

What are the proposed solutions to the regulatory challenges faced by AI in healthcare?

Proposed solutions include more transparent reporting mechanisms for developers to disclose how their models learn and evolve, combined with ongoing and real-time monitoring to ensure accuracy and adaptability.

Why is it important that the training data for AI algorithms is representative of the entire human population?

It is important to prevent data bias and ensure that AI models are trained on diverse and robust data sets to accurately predict outcomes for all populations, not just specific demographics.

What is the potential issue with AI models trained on predominantly one demographic?

AI models trained on predominantly one demographic may not perform well on other populations, leading to potential misdiagnoses or inaccurate predictions.

What functionality is urgently needed in AI models according to the script?

AI models need a functionality that allows them to express uncertainty or provide a best guess based on a skewed training population when dealing with low confidence results.

How can new regulatory frameworks help implement AI in healthcare?

New regulatory frameworks, developed in consultation with stakeholders, can ensure that AI tools are compliant, adaptive, and serve the entire population equally, thus improving healthcare delivery.

Outlines

00:00

🤖 AI's Potential in Revolutionizing Healthcare

Artificial intelligence (AI) is often misconceived as a threat, but it holds immense potential to enhance healthcare. AI enables personalized care delivery, increases hospital efficiency, and broadens healthcare access through precise decision-making tools. It functions by educating computer models with extensive datasets, allowing them to learn and predict outcomes. This technology can analyze patient data, including genetic information, to recommend tailored treatments. AI models can also predict disease susceptibility and treatment responses in different populations, offering a detailed understanding of human health. However, integrating AI into clinical practice is challenged by existing regulatory frameworks not designed for adaptive AI-based software.

05:00

🛠️ Addressing AI's Implementation Challenges in Healthcare

The current regulatory approach to AI in healthcare involves 'freezing' the learning capabilities of AI algorithms before clinical use, which hinders their potential for improvement and accuracy. This can be detrimental as it leads to outdated data influencing diagnoses. Fortunately, new regulatory frameworks are being proposed that emphasize transparent reporting and real-time monitoring to ensure AI models adapt and improve healthcare outcomes. It's crucial that AI training data is representative of diverse populations to prevent bias. An example is a mobile-based AI diagnostic tool for skin cancer detection, which must be trained on varied data to avoid inaccuracies in diagnosing different ethnic groups. Developers must ensure models can acknowledge uncertainty in predictions when data is skewed. Establishing new regulatory frameworks in collaboration with stakeholders is essential for harnessing AI's full potential in healthcare, promoting personalized health advice, and transforming healthcare delivery.

Keywords

💡Artificial Intelligence (AI)

Artificial Intelligence refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. In the context of the video, AI is portrayed as a tool that can revolutionize healthcare by personalizing patient care, making hospitals more efficient, and improving access to healthcare through accurate decision-making tools. The script emphasizes how AI can learn from large datasets to predict patient outcomes and tailor treatments.

💡Personalized Care

Personalized care is an approach to healthcare where treatment is customized to the individual patient's needs, conditions, and genetic makeup. The video discusses how AI can facilitate personalized care by analyzing individual patient data to predict the efficacy of treatments, thus improving healthcare delivery and outcomes.

💡Healthcare Efficiency

Healthcare efficiency refers to the optimal use of resources to maximize the quality of patient care. The script mentions that AI can enhance hospital efficiency by streamlining processes and providing accurate decision-making tools, which can lead to better patient outcomes and reduced costs.

💡Data Sets

Data sets are collections of data points, often used for analysis or training machine learning models. In the video, complex and large data sets are used to educate AI models, allowing them to learn and make decisions or predict outcomes. The script gives the example of using data from blood tests, X-ray images, and genetic information to train AI models for cancer diagnosis.

💡Cancer Diagnosis

Cancer diagnosis is the process of determining whether an individual has cancer and, if so, the type and extent of the disease. The video script discusses the complexity of cancer diagnosis and how AI can assist by consolidating information from various sources to provide accurate predictions of diagnosis and treatment options.

💡Genetic Information

Genetic information refers to the inherited instructions found in an individual's DNA. The video highlights the importance of genetic information in cancer diagnosis, where AI models can use genetic data from tissue biopsies to identify cancer primary sites and inform treatment decisions.

💡Regulation Frameworks

Regulation frameworks are the systems of rules and guidelines that govern a particular activity or industry. The script addresses the challenge of integrating AI into healthcare due to existing regulation frameworks not being designed for AI-based software. It calls for new frameworks that can accommodate the dynamic nature of AI.

💡AI-based Software as a Medical Device

AI-based software as a medical device refers to software programs that use AI to assist in diagnosing, treating, or managing diseases. The video points out that current regulation frameworks are not suited for such software, which is dynamic and learns from data, unlike traditional static medical software.

💡Data Bias

Data bias occurs when a data set is not representative of the entire population, leading to skewed or inaccurate results. The video script discusses the importance of training AI models on diverse and robust data sets to prevent data bias and ensure accurate predictions across different populations.

💡Skin Cancer Detection

Skin cancer detection is the process of identifying skin cancer through visual examination or diagnostic tools. The video gives an example of an AI-based mobile diagnostic software being developed to detect skin cancer from images, emphasizing the need for the model to be trained on diverse data to avoid misdiagnosis.

💡Adaptive Models

Adaptive models are capable of learning and evolving over time to improve their functionality and accuracy. The script contrasts traditional static software with AI software, which is adaptive and can ideally become more intelligent over time. It argues for regulation that allows AI models to continue learning from new data to improve healthcare outcomes.

Highlights

AI can save lives and improve healthcare for millions.

AI personalizes care delivery, makes hospitals efficient, and improves healthcare access.

AI educates computer models using large data sets for decision-making and outcome prediction.

AI models can predict treatment effectiveness based on thousands of patient experiences.

AI helps doctors make informed decisions by learning from similar patient conditions.

AI is being used to deliver care to cancer patients, streamlining diagnosis.

AI models consolidate patient data from blood tests, X-rays, and biopsies for accurate predictions.

AI can identify cancer primary sites, enabling targeted treatments.

AI models can predict healthcare needs and responses to interventions in certain populations.

Regulation frameworks for AI in healthcare are not designed for evolving AI software.

Existing regulations lock AI's learning potential before clinical use.

New regulation frameworks are proposed for transparent reporting and real-time monitoring of AI models.

AI developers must ensure training data represents the entire human population to avoid bias.

AI models need functionality to acknowledge low confidence results due to skewed training data.

Establishing new regulatory frameworks is crucial for successful AI implementation in healthcare.

Improved regulatory frameworks can ensure AI tools are compliant, adaptive, and serve the whole population.

The potential of AI in healthcare can transform personalized health and well-being advice.

Transcripts

Transcriber: Gisele Cristina Ribeiro Reviewer: lisa thompson

Artificial intelligence has often been depicted as villain robots

ready to take over the world.

But I’m here to tell you that AI can actually save lives

and improve health care for millions of patients around the world.

AI is helping us personalize the delivery of care,

make hospitals more efficient,

and improve access to health care by providing accurate decision-making tools.

AI is the process of educating a computer model

using complex and large data sets.

The model learns from this data in a training process

to build its ability to make decisions

or predict outcomes when presented with new data.

We are talking about having access to a computer model that knows,

based on the experience of thousands of other patients,

whether a treatment is likely to work

and what works best for that patient based on their individual conditions.

No two of you in this room or, in fact, anywhere in the world are alike.

But AI models are helping our doctors

learn from patients with similar conditions

or even similar genetic information

and make highly informed decisions

about their diagnosis and their treatment options.

I want to talk about how we are starting to use AI

for delivering care to cancer patients.

Cancer diagnosis can be immensely complicated,

both for the doctors

in making decisions about diagnosing a primary or secondary cancer,

as well as for the patients, in understanding the risks

and success rates of the treatment options.

But we are developing AI models that can help streamline this process

by taking information from a number of sources.

This involves feeding an AI model data from the patient’s blood tests,

X-ray images of the suspected lesions,

as well as genetic information from a tissue biopsy.

The trained AI model can rapidly consolidate this information

and provide highly accurate predictions of the patient's diagnosis,

treatment options most likely to succeed, as well as the prognosis.

Let’s talk about Peter, who is a cancer patient.

He’s gone through comprehensive clinical assessment, imaging

and various other diagnostic workups,

but not even the best doctors in town can tell him where his cancer primary site is,

meaning he can’t get a treatment specific for his cancer

and his chances of surviving another five years is less than ten percent.

But our team right here in Brisbane

has developed a tool using AI and patients’ genetic information

that can accurately identify the cancer primary site of Peter,

and empower doctors

to give Peter a treatment that we know is going to work for him.

These type of models can be expanded exponentially

to predict accurate health care.

This means using an AI model

to understand whether a certain population is more susceptible to a certain disease

and whether they would respond more favorably

to certain health care interventions.

AI is giving us the ability to have a much more refined

and detailed understanding of human health than we’ve ever had before.

But there is a catch to the immense promise of AI

being implemented into routine clinical practice.

Our existing regulation frameworks aren’t designed for AI software

intended for diagnosing, treating or managing the disease,

also known as AI-based software as a medical device.

They are designed for physical medical devices, like surgical implants,

or most software that have the same output

every time that the patient or clinicians are using them.

Traditional software are static,

in a sense that the developers release a version of a software

and, no matter how many times you use it,

it would always have the same output for the same data.

On the other hand,

AI software behaves completely differently to most software in health care

because of the intrinsic ability to learn and evolve over time,

ideally becoming more intelligent

as suited to the environment that they’re being used at.

Our existing regulation frameworks

rely on the static and reproducible nature of this software

to prove that they are safe

to be implemented into routine clinical practice.

So, our regulatory authorities’ solution has been to lock the learning potential

of these algorithms before they are implemented into clinical practice.

This means that the model can no longer learn from its environment and new data,

which limits its potential to improve its functionality or its accuracy,

you know, the whole point of AI.

And, at times, this can even be harmful for the patients

because the AI model is no longer trained on the most up-to-date data

and can potentially lead to a wrong diagnosis.

But the good news is

that there are emerging regulation frameworks being proposed

that, if implemented right, can be a game changer.

Our regulatory authorities are proposing using more transparent reporting mechanism

so that the developers can disclose how their models would learn

and evolve over time.

And this will be combined with ongoing and real-time monitoring

to make sure that the predicted changes actually occur

and that the software is adaptive to make much more accurate predictions

and improve health care outcomes.

We also need to make sure

that the training data used for these algorithms

are representative of the entire human population.

Let’s look at a mobile-based diagnostic software

that we are developing right here in Brisbane

that uses AI to detect skin cancer

from the images that you’ve taken on your iPhone.

If this model has been trained on a predominantly Caucasian population,

how well do you think it would do

on an African American or an Asian patient?

Our AI developers have a huge responsibility

to make sure that data bias doesn’t exist

and that their models are trained

on diverse and robust data sets, representative of the entire population,

you know, not just white males.

But at times, we understand that this is not entirely possible.

Skin cancer does, in fact,

disproportionately affect the Caucasian population

because of the genetic differences,

and, as a result, there are much larger data sets available for those patients.

But this means that we need to build in a functionality in our AI models

that, for low confidence results, for an Asian patient, for example,

the model is capable of saying “I don’t know”

or that “This is my best guess based on a skewed training population.”

But, unfortunately, this functionality doesn’t exist yet,

and it’s urgently needed to be mandated by our regulators.

To successfully implement AI in health care,

we need to establish new regulatory frameworks

in consultation with AI developers, health care practitioners,

policy advisers,

as well as the patients themselves, to bring the best out of AI.

Improve the regulatory frameworks

can make sure that diverse and robust tools are developed

that are compliant and adaptive

and can serve the whole population equally.

If we get this right, we can transform the delivery of health care

where we are promoting personalized health and well-being advice.

I’m excited to be at the forefront of translating this amazing technology

into health care

and use this to help millions of lives around the world.

(Applause)