Artificial intelligence will radically improve healthcare, but only if managed carefully.

More important than speeding up the widespread use of artificial intelligence (AI) in American health care, is making sure we do it right. To unlock the greatest positive effects of innovation, ensuring integrity and transparency must be a top priority. This can be accomplished by applying the principles that guide medical research, including respect for the human person, maximization of benefits and avoidance of harm to patients, only benefit sharing, meaningful informed consent and Protection of confidential patient information.

The emergence of artificial intelligence is reminiscent of the Great Gold Rush, a unique time bursting with limitless potential that is filled with uncertainty, speculation, and unpredictable outcomes. Advances in AI are leading to a true transformation of medicine that can help reduce existing burdens and inefficiencies while simultaneously improving patient care and experience. Examples range from ambient voice transcription tools that enable doctors or nurses to spend more time with their patients to diagnostic devices that detect diabetic retinopathy or colon polyps, the list grows daily. Its applications are almost limitless. A new revolution has come.

This technology has revolutionized the healthcare sector, but it still has a long way to go for widespread implementation. It remains to be seen how medical professionals and patients will interact with and use artificial intelligence. Unfortunately, the potential for harm has already been demonstrated by examples of substantial algorithmic bias and the use of AI to deny patients permission to care. Experts use the term Human in the Loop (HITL) to describe the required human involvement within an automated process system. However, this is insufficient because we need progressive machine learning to be not just one dimension of a system, but at the top of the hierarchy. The last line repeats: Humans must live at the top of the hierarchy. We need to control AI, not the other way around.

The complexity of artificial intelligence will require significant bandwidth to properly monitor its application and put in place sensible backbones that enable innovation while protecting patients and other key stakeholders. The scale and scope of this initiative is far beyond what the federal government alone can do. Unlike top-down approaches in other parts of the world, we must use public-private partnerships to develop these guidelines and guidelines and ensure that what is produced is reliable and valuable. Is. This can be achieved, in part, by creating independent assurance laboratories that evaluate AI models and their applications using generally accepted principles. We need multiple chickens to guard the hen house.

Avoiding the same mistakes that now hinder the integration of mature technologies, such as electronic health records, is paramount. National standards are important for establishing health AI best practices for the use of emerging innovations, and adoption of these benchmarks should be as close to end beneficiaries as possible. The federal authority has an important role to play here, as a convener and enabler in creating these standards. However, their implementation should be deferred as much as possible to local governance at the health system level, with federal authorities intervening when necessary. Progress will not be free, but we must learn from past mistakes.

In our quest to mainstream artificial intelligence, ethical considerations must remain paramount. Patients from rural or low-income groups should have access to the benefits of this technology. Further, it is important that the AI ​​used on or by these communities is as reliable as that used by higher health systems. Just as access to healthcare is not a guarantee of quality, access to artificial intelligence systems will not guarantee the capability or reliability of what is available.

Reducing the burden on clinicians, improving patient health and experience, and introducing new life-saving technologies into the ever-expanding world of healthcare is an exciting endeavor. Transcending these unknowns in a way that avoids avoidable risks will allow human intelligence to harness the power of infinite computation to create better and more affordable care. Practitioners and patients alike expect the powerful capabilities and practical benefits of artificial intelligence in healthcare delivery. It is important to ensure that its impending and explosive entry into care settings is handled judiciously and strategically to maximize its positive effects for all.

Greg Murphy, MD, a practicing urologist, represents the Third District of North Carolina. Michael Pancina, PhD, serves as Chief Data Scientist at Duke Health and Professor of Biostatistics and Bioinformatics at Duke University School of Medicine.