Generative AI in healthcare: Examples, benefits, use cases

The wild popularity of OpenAI’s ChatGPT has sparked a race to incorporate generative AI into applications used in industries. Healthcare is among those leading the charge.

Generative AI in healthcare could help unlock a piece of the unrealized $1 trillion in potential improvement in the industry by automating work that’s prone to errors, providing a wealth of data to clinicians in seconds, and modernizing health infrastructure, according to McKinsey & Company.

Examples of generative AI in healthcare

Although healthcare businesses have used AI technology for years — adverse-event prediction, operating-room scheduling optimization, and connecting patient data to drive better outcomes are three examples — AI in healthcare promises even deeper transformation for the industry.

Here are some recent examples of AI in healthcare:

• Amazon Web Services (AWS) in July announced a service called AWS HealthScribe, which uses speech recognition and generative AI to save clinicians time by generating clinical documentation.
• Google is testing medical chatbot technology called Med-PaLM 2 at the May Clinic and other hospitals, according to the Wall Street Journal. Based on the large language model (LLM) technology underlying Google’s own conversational generative AI chatbot Bard, Med-PaLM 2 aims to answer medical questions more accurately and safely.
• Microsoft, one of the major investors in OpenAI, the company behind ChatGPT, is partnering with Epic Systems to integrate generative AI technology into its electronic health records (EHRs). Microsoft’s Nuance Communications subsidiary also announced a fully automated clinical documentation application that combines conversational and ambient AI with GPT-4, the latest version of the LLM powering ChatGPT.

83% of the most innovative healthcare companies in the world run SAP solutions.

Harnessing data in one of the most data-intense industries

To witness such traction in healthcare, a conservative industry that is notoriously one of the last to embrace new technologies, is remarkable. That’s particularly true when we consider how strict privacy regulations like the Health Insurance Portability Accountability Act (HIPAA) tend to deter data sharing.

Yet healthcare is also one of the most data-intense industries.

The average hospital is said to produce about 50 petabytes of data every year, which adds up to approximately 12.5 trillion digital copies of the King James version of the Bible. What’s more, the volume of data generated in healthcare is reportedly increasing by 47% per year, a significant clip for any industry.

All that data must be logged by someone, which takes considerable time. A lot of it could be extremely useful for improving the efficiency of healthcare organizations and for providing timely medical information and advice to patients in need.

However, making good use of that data is all but impossible because there’s far too much of it for human beings and older technology to handle.

Here’s where AI comes into play. By relying on deep-learning algorithms to create new text, audio, code, and other content, it can be optimized for privacy and then lasso huge volumes of unstructured medical information to save time and money while unlocking endless business and clinical possibilities.

Personalized medicine, faster diagnoses, and more

Healthcare organizations see this potential, which is one reason why 64.8% of them are exploring generative AI scenarios and 34.9% are already investing in them, according to IDC Health Insights Analyst Lynne Dunbrack.

“There is demand for technology to address key priorities – such as enhancing patient experience, improving population health, and reducing costs,” Dunbrack says.

More specifically, analysts say generative AI could become integral for addressing a host of common needs in the healthcare industry, including:

1. Personalized medicine: Generative AI can aggregate and analyze huge volumes of patient data to deliver tailored medication and therapy recommendations to enhance treatment.
2. Medical imaging analysis: LLMs also excel at interpreting medical images, such as MRIs, CT scans, and X-rays. Automating image analysis could allow clinicians to detect problems more quickly and accurately, improving diagnosis and overall care.
3. Clinical note taking: An array of tools like AWS’s HealthScribe will likely seek to solve the age-old problem of doctors being too limited on time to take proper, detailed notes on patient visits and then having to enter them into EHRs. AI tools could alleviate that hassle by using speech recognition and deep learning to automate these processes, quickly and efficiently.
4. Drug discovery and development: Generative AI can also be used to hasten drug discovery by simulating molecular interactions and predicting possible drug candidates. It’s also being used to speed up regulatory drug approvals. For example, Insilico Medicine says it not only used its own AI platform to discover a treatment for idiopathic pulmonary fibrosis, a rate respiratory disease, but has applied it to each step of the preclinical drug discovery process. Insilico says doing this through traditional methods would normally be an expense of more than $400 million, but it expects to accomplish this at about a tenth of that cost with generative AI.
5. Virtual health assistants: Many of the first generative AI chatbots for healthcare will focus on providing faster and better answers to patient questions more efficiently 24 hours a day. UNC Health, for example, is tapping Epic’s generative AI tools to help overtaxed clinicians cope with the crush of messages coming their way.
6. Streamlining insurance claims: Generative AI also helps streamline medical insurance claims from patients (or pre-authorization requests from doctor’s offices) by automating the extraction and processing of patient records, thus alleviating pressure on call center personnel. Additionally, it can help speed consideration of claims while identifying potentially fraudulent submissions by detecting suspicious patterns and anomalies.

Generative AI as a technology is still in its infancy, despite all the early hype, so it’s impossible to predict whether the current focus on healthcare will persist. But one thing is clear: with so many billions of dollars being injected into the space, the trend won’t lag anytime soon.