Many in the field of health care today have great hopes of artificial intelligence to do everything from the defeat of doctors to the progress of medical research. The challenge with artificial intelligence in health care is not excitement, there is a lot.
Instead, they are the main challenge of IT leaders is clarity – the vision of artificial intelligence clearly what it can actually accomplish. Where is artificial intelligence already transferred to investment? What is safe to publish now? How can you manage the risks, governance and long -term value without being swept into noise?
Dr. Justin Norddin is a doctor and founder of Stanford and CEO of Al -Mudhala Company, who sells the infrastructure of the Obstetric IQ in health care designed to provide technology, training and support to hospitals and health systems to start artificial intelligence and expand its scope safely through institutions.
Focus on the word caution
It warns of information technology leaders in healthcare organizations to determine the return on investment, safety, risks, governance and long -term value when approaching artificial intelligence systems and tools. It focuses on the word caution.
“We are now almost For two and a half years in the ChatGPT versionAnd although the tanner about artificial intelligence in the field of health care is higher than ever, it is time to ask, “Where is the return on real investment?” “Despite all the excitement, even the most reliable state of use – the surrounding documents – did not provide consistent financial returns through the provided groups. Some doctors love them, but adoption is still limited and uneven.
“Meanwhile, the dramatic titles around AI excel over doctors in diagnostic tasks, but these titles are exacerbated, but these cases are clinical use are what will determine the impact of artificial intelligence in the short term in health care.”
He claimed that the true value of artificial intelligence today is in health care.
“This is the place where we started to see the return on investment,” he said. “Artificial intelligence can now open visions of unconfirmed data-the largest part of what health care produces. It may not seem tasks like reporting quality, improving the work course function, and simplifying the patient’s awareness of cheerful, but necessary and taking a long time.
“Artificial intelligence is finally able to automate what was buried in PDF files, fax and clinical notes.” “These improvements behind the scenes may look small individually, but together they represent a major and developed effect.”
Ideas from people closest to work
Many in health information technology and health care are still looking for one “killer application” to change everything. But the real shift will come from hundreds – or thousands – from small and practical use cases in daily work – and the best ideas will not come from top to bottom but from people closest to work.
“Doctors and nurses already use artificial intelligence Unofficially, on personal devices or through solutions, “he pointed out.” This tells us two things: there is a request, and there is a danger. The path forward is clear – we need to bring artificial intelligence over the table. Make it safe and compatible with HIPAA and can be accessed so that we can convert this quiet revolution into permanent progress at the system level. “
On another front, when it comes to spreading artificial intelligence safely in health care today, Norddin said it is very important to start admitting what is not safe – because this is the place where many organizations are still open, whether they know that or not. One of the most urgent concerns is the employees who use personal intelligence accounts to process sensitive patient information – which is more common than many.
“He spoke to leaders throughout the country and you will hear everything from, we know that it is happening, but we look in the other direction” to “we will deal with it if it becomes a problem.” “Some even work under a calm policy” do not ask, do not tell “, but none of these are long -term strategies. We have seen this before with the Pixel and Google Track from Google as they violate privacy violations as soon as they are ignored.
“Another field that requires caution is the tools of general artificial intelligence.” “While illustrations can be impressive, these systems are vulnerable to” fracture of protection. “We have seen artificial intelligence tools exploited to produce inappropriate, harmful or even dangerous content, and often go beyond the intended protection of the systems.
Watch out of the open internet
He added that the risks grow dramatically when these models are connected to the open Internet.
“Poor actors can cultivate harmful online content designed to influence artificial intelligence behavior, creating serious threats to cybersecurity,” he said. “At best, this leads to a headache of public relations. In the worst case, it can lead to data violations or ransom attacks that prevent complete systems.
“The safest path begins forward Amnesty International’s internal publication In the safe environments compatible with HIPAA with humanitarian systems in the episode, he said: “This makes sure that data is used safely, and people still sign the measures taken by artificial intelligence systems. Early artificial intelligence applications must focus on operational areas such as simplifying the supervisor’s tasks, improving workflow and reducing friction – areas that provide investment return without an introduction of a clinical risk. The goal is not to avoid artificial intelligence – It is wisely using it, building value and confidence with each step.
Norden also provides warning advice when it comes to managing risks, governance and long -term value with artificial intelligence without getting a cliff in noise, describing this one of the biggest challenges in health care today.
“There is a natural hesitation, rightly, given the amount of what is still unknown,” he said. “For this reason, many health systems are stuck in cautious mode – the launch of the pilots, and they are engaged in internal side projects and experimenting without a clear path forward.
“The shift from” this seems promising “to” this is safe and developed “begins with clear leadership and the direction about the tools needed to use it, and how we should measure success before we start.” “The difficult thing is without a clear directive of our workforce now, people turn into external public tools and under the use of the table. We must start with options compatible with HIPAA for the place where our workforce should reach these tools.”
More than safety
But safety alone is not enough.
“Often, the most secure tools are more effective or less useful, which causes people to return to public options.” “We need to make the inner tools safe and really more valuable. This means including artificial intelligence in a real workflow and enriching them with internal data, so not only compatible, but it is also indispensable.
“With the expansion of use throughout the organization, governance should also expand.” “This includes following use, scrutinizing reactions and educating users-not to buy them, but to direct safe and responsible use. If someone tries to use AI for highly dangerous tasks such as pharmaceutical doses, we need systems in place and correct this behavior early.”
He added that the long -term value in the end comes from building a repetitive processable process.
“This means that the organized pilots, performance thresholds, and infrastructure that help the governance teams to track and develop what succeeds,” he said. “Through strong tools, smart policies and clear priorities of driving, we can move previous experiences and to a sustainable transformation at the system level.”
Avoid common mistakes
How can hospitals and health systems avoid common mistakes that give progress? Nurden said in a variety of roads.
“At the present time, when we talk to healthcare leaders, we see most of the artificial intelligence strategies located in one of four buckets-waiting for an EHR seller to put something, or prohibit tools like ChatGPT, or buy a points system like surrounding documents, or try to build everything at home,” Nordin noticed. “All of these methods have some logic behind, but they often miss the largest image.
“What is really required is a clear and common vision throughout the institution to come, It will change the way in which health care worksHe continued: “Without this purchase, it ends with the difference in silos, not sure of focus and progress.”
There is another common dilemma that tries to do a lot at the same time.
“We all have seen systems chasing dozens of pilots with different sellers, spreading their time and delicate resources,” Norden said. “The result is not enough to draw anywhere to make a meaningful effect. What works better is choosing a few high priority areas where AI can make a clear and immediate difference and invest in those who have real support and support support.
He added, “It is related to creating a fewer betting and more intelligent bets and giving these teams the tools, data and clarity they need to achieve success,” he added. “This concentrated approach builds momentum and makes it easy to expand the scope of what works.”
Do not forget people
Finally, Norddin said that one cannot talk about avoiding mistakes without talking about people.
“Most of our employees already use artificial intelligence tools in their personal lives, and they are increasingly bringing them to work.” “If we ignore it or try to close it, we miss a great opportunity. What we have to do is to tend to it by giving them safe and safe tools to experience and teach them how to use artificial intelligence effectively and responsibly.
He concluded that “education and training cannot be once, and it should be a continuous part of how to support our teams.” “The future of artificial intelligence in health care is not only related to technology – but rather to enable our people to use it well. When driving brings everyone, that is when a real transformation occurs.”
Follow Bill Hit coverage on LinkedIn: Bill Seuiki
Email him: bsiwicki@himss.org
Healthcare is Hosz News.
Watch now: seize the head of Ai Brass Ring – and work with the upper copper
