[ad_1]
Healthcare is fertile ground for artificial intelligence for several key reasons. It’s large and complex, with millions of lives and futures at stake, so the opportunities are endless, but there’s also less room for error than in other fields.
To help professionals and policymakers understand how artificial intelligence is evolving, where it is heading in healthcare, and the milestones it has reached along the way, the National Institute for Healthcare Management recently hosted a network on the topic seminar. Panelists Dr. Michael Matheny of Vanderbilt University Medical Center, Svetlana Bender of GuideWell and Florida Blue, and I. Glenn Cohen of Harvard University joined NIHCM’s Kathryn Santoro on the challenges and opportunities presented by this rapidly evolving and evolving technology. Lectures were given.
Get ready for artificial intelligence
Matheny noted that healthcare can benefit from artificial intelligence precisely because of the vast amounts of data generated and available in the field. Clinicians need to process “a huge amount of scientific information,” Matheny said, adding that artificial intelligence could be deployed to streamline the process of clinical trial analysis and practice guidelines. “We need help managing all this information,” Matheny said. Masini pointed out that artificial intelligence can also be applied to areas such as clinician note creation and clinical decision support.
Matheny cited a recent finding that pulse oximeters provide inaccurate readings for some patients of color. Matheny said the training materials used by these devices are based primarily on the white population and therefore are not representative of people from other racial groups. This is just one application where artificial intelligence can be deployed to improve accuracy and eliminate bias.
“It’s really important to understand what you want to change or what needs to be fixed and then evaluate it in the context of the workflow, the stakeholders, the end users that will be affected (patients, caregivers),” Matheny said, adding says that, in some cases, artificial intelligence may actually no is the answer.
Overcome obstacles
Bender noted that artificial intelligence can be seen as a bridge between minds and machines, and is used in ways people may not even realize, from asking Google Maps to find the fastest driving route to choosing a movie or show suggested by Netflix. “It’s been around for decades, but it’s only recently…that it’s captured the public’s imagination,” Bender said.
Bender said using artificial intelligence in health care has the potential to save $200 billion to $360 billion annually. Wearable health devices, symptom checkers, medical imaging, accelerated drug discovery, medical claims approval, and Medicaid fraud detection are all areas where artificial intelligence can help. Still, adoption in the health sector remains low, which is “a big problem,” Bender said. “In fact, less than 5% of healthcare organizations are using artificial intelligence as of 2022, a number that really lags behind many other industries.”
Bender said slow adoption could be related to three important behavioral factors, and he suggested ways to address them:
- fear of change This problem can be mitigated by involving people from all parts of the healthcare organization, including legal, information technology, marketing, and business operations.
- fear of unknown This can be addressed by upskilling employees, asking for their input and explaining that AI applications are intended to enhance their jobs, not necessarily replace human jobs.
- fear of algorithms This problem can be overcome by focusing on good governance, proper controls, transparency, human oversight and elimination of bias.
GuideWell has used AI technology to personalize patient care, streamline prior authorization approvals and create apps like chat tools for staff, Bender said. “[S]Bender said “structure, training, ethics, technology and partnerships” are key pillars in adopting artificial intelligence.
Make artificial intelligence fair and effective
Cohen provides insights into the legal and ethical considerations of artificial intelligence in health care. When an organization builds a model, problems can arise at any stage of the life cycle, so it’s important to understand the process, Cohen said.
- At the first stage During the development process, organizations should consider factors such as data sources, removal of personal identifiers, and representativeness of the population.
- second stage Involves constructing and validating models. Questions have been raised about standards, the reliability of verifiers, and the balance between intellectual property protection and the need for transparency.
- The third phase Involves testing in real-world settings, understanding accountability, consent and regulatory controls, and deciding how to inform patients about the use of artificial intelligence in healthcare.
- in the fourth stagedisseminate and use functional models for the benefit of patients and monitor their fair use and commercial viability.
Cohen noted that while medical AI may make mistakes, be biased, or fail to explain its output in some cases, skeptics should remember that the same goes for the clinicians providing care. As far as artificial intelligence is concerned, this process will only improve over time. When it comes to accountability, Cohen said, AI can serve as a “validation tool” to support existing decision-making methods. For example, it may help notify clinicians when a particular type of care is inappropriate.
Artificial intelligence has the potential to truly become the standard of care one day, but it could ultimately pose greater risks, Cohen said no Use it in healthcare settings. We’re not there yet—but it may be the path we’re on, Cohen says.
________________________________________
Take advantage of SmartBrief’s free email newsletter For healthcare CIOs and leaders, among companies Over 250 industry-focused publications.
[ad_2]
Source link