The use of AI in healthcare has several advantages, including reducing burnout among doctors and improving clinical trials.

• Generative AI is now prioritizing efficiency in health care, while traditional AI has focused on making it safer and better.
• Medical professionals experienced less burden and burnout when using AI to generate draft replies to patient inbox messages, despite spending the same amount of time on the task.
• AI-powered tools are emerging to streamline the process of matching participants to clinical trials, speeding up drug development, and simplifying the translation of documents for non-English speaking patients and trial participants.

While traditional artificial intelligence has primarily focused on improving healthcare outcomes, recent advancements in generative AI have shifted the focus to efficiency.

Nvidia, a hardware and chip company, has been working to optimize the health care space for 15 years. Kimberly Powell, Nvidia's vice president of health care, and her team build domain-specific applications for health care, including in the realm of imaging, computing, genomics and drug discovery, under the umbrella of the "Clara" suite.

Powell stated that the process involves connecting small applications to provide a valuable service to the market.

The healthcare industry is a large data industry, Powell notes, and as such, it is heavily regulated and requires careful handling when bringing it to market.

"Dr. Josh Fessel, director of the office of translational medicine at the National Institutes of Health, stated that while some may believe that they are late to the game, he is not convinced of this notion. He emphasized the importance of being cautious when dealing with human beings and the need to prioritize privacy, security, and transparency."

Fessel's specialty is translational medicine, which transforms promising ideas into tangible solutions that benefit people. Currently, AI is the driving force behind this quest.

While AI is being used to improve contact centers, modernize code, and reduce medical burnout, Dr. Kaveh Safavi, senior managing director for Accenture's global health care business, emphasizes the importance of medical professionals learning to communicate their findings in the exam room. "This is a reality that must be accepted," he said. "The technology requires the human to adapt in order to fully benefit."

Fessel argues that while a March study showed that AI-generated draft replies to patient inbox messages reduced burnout scores in medical professionals, it did not decrease the amount of time they spent on this task. However, time is not the only factor that matters, Fessel emphasizes.

AI to address nursing shortage

New technologies are expected to be deployed within the next year to address the global health emergency of the nursing shortage, according to Safavi. Meanwhile, AI-enabled solutions are being developed to efficiently match potential participants to clinical trials, expedite drug development, and translate documents for non-English speaking patients and trial participants.

Despite the challenges, there are still issues to address. For instance, a study conducted by the Clinical & Translational Science Award (CTSA) Program for the Mount Sinai Health System in October revealed that predictive models that use health record data to determine patient outcomes can influence the real-world treatments that providers give those patients, ultimately reducing the accuracy of the technology's own predictions. In other words, if the algorithm performs as intended, it will alter the data - but then it operates on data that is different from what it learned, ultimately reducing its effectiveness. "It creates its own world, essentially," said Fessel. "This raises the question: What does continuing medical education for an algorithm entail? We don't know yet."

Fessel suggests that a team approach across institutions is necessary to address knowledge gaps. He emphasizes the importance of sharing what we're learning. Having a chief AI officer in a health institution can be helpful, but only if they are empowered to bring in other brains and resources.

Nvidia collaborates with various organizations to implement "microservices," which integrate into existing applications. This aids in navigating regulatory changes, such as those related to software as a medical device (SaMD) per the U.S. Food & Drug Administration. As an example, Nvidia partnered with Abridge on its first application, which streamlines medical summaries within the Epic electronic health record system.

Nvidia is partnering with Medtronic and the Novo Nordisk foundation to enhance computer vision technology in colonoscopies, detecting 50% more potentially cancer-causing polyps. Additionally, they are collaborating to establish a national center for AI innovation in Denmark, which will house one of the world's most powerful AI supercomputers.

Safavi states that provider organizations are currently prioritizing preparing for generative AI. This involves ensuring their technology is ready for cloud-native tools that require access to data.

A human is the 'last mile'

Safavi emphasized the importance of maintaining a responsible AI posture that safeguards privacy and intellectual property while discouraging the use of technology for diagnosis. He emphasized that humans should be the final decision-makers in judgment.

Safavi expressed his concern that the lack of policies against technological diagnosis in the health care industry could lead to negative consequences. He emphasized the importance of being proactive in establishing boundaries to prevent an overly generalized regulatory framework that would not benefit anyone.

The European Union adopted the Artificial Intelligence Act in March, which includes provisions on AI safeguards, biometric identification systems, social scoring, and the right to complaint and transparency. Safavi, who has worked in over 25 countries over the past 15 years, predicts that any regulatory system the U.S. adopts will likely mirror that of the E.U., but we are not there yet.

Despite the numerous advancements in medicine, there remains a significant amount of uncertainty regarding the development of various health conditions and the impact of the environment. According to Fessel, acknowledging the existence of "black boxes" in medicine is crucial to reevaluating fundamental concepts and advancing the field. He believes that this redefinition of healthcare has the potential to be transformative.