Close

See How Your Peers Are Moving Forward in the Cloud

New research from CDW can help you build on your success and take the next step.

Jul 31 2023
Patient-Centered Care

Generative AI in Healthcare: 5 Questions Surrounding It

Two healthcare and life sciences attorneys dive deeper into what the recent growth of generative artificial intelligence tools means for providers and patients.

Governance over the use of artificial intelligence in regulated environments such as healthcare has been developing in an almost arbitrary manner.

The U.S. Food and Drug Administration has taken the most definitive steps in its publication of the 2021 Artificial Intelligence and Machine Learning Software as a Medical Device Action Plan, which outlined the establishment of “practical oversight” of AI or ML-based Software as a Medical Device. However, the first draft guidance on proposing an approach to support iterative improvement wasn’t released for comment until spring 2023.

Other government entities have also issued frameworks, guidance and principles, including the Federal Trade Commission, the National Institute of Standards and Technology, and the White House Office of Science and Technology Policy. Common themes of data bias, transparency and accountability continue to emerge, but without clear guardrails or even practical guidance.

Here are five questions for healthcare organizations as the industry enters a new era of AI deployment.

Click the banner to discover how health IT solutions can help create an integrated care experience.

1. What Is Generative AI, and What Are Some Examples of It?

Generative AI is a type of AI system that uses data to generate new content, which can include, for example, text for patient electronic health records, enhanced medical imaging, or synthetic data, to train itself or others for drug discovery and development.

Popular chatbots and text-to-art generators include ChatGPT and DALL-E 2. Generative AI has been around for years but has exploded in popularity due to recent advancements in ML algorithms that have expanded the capabilities and potential use cases of generative AI, including in healthcare. There are many other types of AI being used in healthcare, too.

2. What Makes Generative AI Appealing to Healthcare?

Healthcare is all about Big Data, and generative AI has an uncanny ability to process enormous amounts of data that would otherwise take a person months or years to study. In a healthcare context, diagnostic, research and other generative AI tools will be capable of drawing from a far deeper well of knowledge, as well as reviewing and analyzing more medical literature, studies and clinical outcomes, than any single person could within a lifetime. 

The power and reach of Big Data training sets was demonstrated during the COVID-19 pandemic, when researchers and doctors used AI and ML algorithms to quickly review constantly changing data, assess geographic hotspots, track spread and diagnose COVID-19 pneumonia versus common pneumonia. Even developers cannot predict the “capability overhang” or hidden capacities of the undiscovered potential of AI algorithms in healthcare.

FOC Sidebar

 

3. How Does Generative AI Differ from Other AI/ML Solutions?

Generative AI is distinct from other AI solutions that simply find the requested information from many sources. In contrast, generative AI creates the content based on predicted patterns gleaned from the training data sets.

But with great power also comes great responsibility. The U.S. Department of Health and Human Services’ 2021 Trustworthy AI Playbook was prescient in identifying six primary concerns for AI use in healthcare: data privacy, impartiality and bias, transparency in algorithms, responsibility and accountability, data safety and security, and robust and reliable results.

“Hallucinations,” or outputs that are fabricated and often downright false, have already made news and made companies view generative AI more cautiously. People took notice when two lawyers faced sanctions for using ChatGPT to write a brief — which relied on case law that simply did not exist. 

In healthcare, the stakes are even higher: Hallucinations can be inefficacious at best and deadly at worst. 

LEARN MORE: What is digital health and how is it evolving?

4. What are Potential Use Cases for Generative AI in Healthcare?

To date, generative AI has been used to help automate the mundane tasks of medical record documentation, generation of patient instructions, and billing and coding. Also, in 2022, the FDA approved AI medical devices that largely fell into broad categories of AI-assisted medical imaging and diagnostic tools and personalized health plan software.

As AI tools improve and trust grows deeper, the sphere of influence will move from the routine tasks that revolve around healthcare to the core functions that constitute healthcare: discovering and developing drugs, preventing unnecessary surgery and creating synthetic “patients” for training purposes.

While generative AI tools already are playing roles in such use cases, their roles will become more typical and even expected as models become even more sophisticated. 

EXPLORE: How data and artificial intelligence can improve outcomes.

5. How Can Generative AI be Integrated into a Healthcare Environment?

In the absence of a clear legal and ethical framework that promotes the complete scope of the benefits of generative AI across the healthcare ecosystem, the full potential of the technology cannot be attained.

AI healthcare applications today are cautiously serving in the administrative and supportive roles of the delivery system. But the true potential lies in the shift of technological tools from assistive to primary instruments of medical care. To integrate those tools into the healthcare environment, there must be a framework that addresses hard ethical and legal questions. Who will be practicing medicine: the doctor or the machine? What role should governments play in ensuring safety, impartiality and accountability?

Perhaps this lack of clarity is a result of the far-reaching and ever-changing nature of the technology, which does not lend itself to a single, comprehensive set of laws specific to the technology itself. Instead, we should expect to see those laws and regulations that have applied to the traditional healthcare field adapt and change to address the novel issues raised by generative AI technology. The already complex and highly regulated healthcare environment will need to be re-ordered with intention and creativity to advance at the speed of data.

(digital composite): jittawit.2 (background), SDI Productions (woman), AndreyPopov (screens), SDI Productions (hands)/Getty Images