The American Medical Association (AMA) recently released a report on augmented intelligence (AI) in health care. Clinical informaticist and senior fellow for Innovation at OSF HealthCare, Jonathan Handler, MD contributed to the report.
The report aims to create a shared language for AI in health care and emphasizes the AMA’s stated commitment “to ensuring the physician voice is leading health care AI forward.” It covers both current and future uses of AI, including the role of generative AI in health care diagnosis and automation. It also discusses the potential risks that doctors need to be aware of, such as AI bias in health care and privacy concerns. The report provides key questions to help doctors navigate the use of AI in their practices responsibly.
“I think the world recognizes that artificial intelligence is a really important advancement for medicine, one that’s disruptive and has the potential for both benefit and harm. We want to make sure that it creates the greatest benefit possible for everyone, most importantly, for the patients that we’re serving,” says Dr. Handler.
Significance of the AMA report
The topic of AI is trending given the recent introduction of generative AI and large language models like ChatGPT to the general public.
“These things are now more accessible to non-technical people so they can actually use it and see how impressive the results can be,” says Dr. Handler. “We can all see the potential, and when we use it and don’t get a good result, we can also all see the potential downsides.”
AI in health care is not new, but its use is continually expanding, redefining the delivery of care. The report notes that AI has been used in health care since the mid-20th century, mostly in the form of rules-based algorithms that rely on human input to create the desired output.
“Many things we use commonly in medicine today depend upon or are enhanced by algorithms. Soon after becoming comfortable with technology, many of us forget that it’s using algorithms. The pulse oximeter that provides information about a patient’s oxygen status, the ECG and cardiac monitors, the MRI and CT scan machines – all these and many more produce data that are processed and made useful by algorithms,” explains Dr. Handler.
Augmented vs. artificial intelligence
The need for transparency comes as generative AI in health care advances and has the potential to play a bigger role in the clinical decision-making process.
The report explains the difference between artificial intelligence and augmented intelligence. The AMA prefers the term augmented intelligence “to reflect its perspective that artificial intelligence tools and services support rather than explicitly replace human decision-making.”
AI in health care is already playing a huge role in making some health care delivery tasks easier. Many health care systems already use generative AI for tasks, such as clinicians using real-time voice-to-text transcription when creating their clinical notes. This creates more time for the clinician-patient interaction and a better overall patient experience.
When considering decision support capabilities, like an AI suggesting a medical diagnosis or an AI translation of medical data, the risks of AI in health care become more obvious.
Human interaction with the results of generative AI in health care is key. One of the cons of AI in health care is being too trusting of the data it generates.
“We need to educate people to be careful and not to put more trust in AI models than is warranted. We must help improve AI so it can better meet our needs and those of our patients. Ultimately, we need reliable, trustworthy, safe, fair and effective AI,” describes Dr. Handler. “Right now, in healthcare we have to lean a little bit more on people because not all of our processes and technologies are fully optimized. However, I think our people, processes and technologies are evolving quickly and I have high hopes for the future.”
The impact of AI at OSF
Want to collaborate on AI?
The Advanced Analytics team was created in 2013 as part of the OSF Innovation ecosystem to use complex data modeling techniques to serve care delivery teams and their patients. With experts in data science, automation and artificial intelligence, Advanced Analytics collaborates to understand problems, define focus questions, prepare data, build effective models and implement solutions to support clinical actions.
In 2021, to complement the Advanced Analytics team, OSF created what later became the Clinical Intelligence and Advanced Data Lab (CIADL), led by Dr. Handler and Roopa Foulger. The goal of CIADL is to create transformative innovations that help every patient receive a correct diagnosis, optimal therapy and achieve their best possible health in the shortest possible time.
At OSF, the commitment to innovation includes the responsible use of AI. OSF has formed a multidisciplinary AI governance and leadership committee that supports the use of AI to help OSF provide health care with the greatest care and love.
“OSF wants to make sure that we’re not benefiting one group while harming another. We want to use AI to benefit everyone,” says Dr. Handler.
That benefit is clear across OSF in many ways.
- The NeuroHealth Lab is using AI technology to help clinicians detect and diagnose neurological conditions.
- OSF uses and has invested in IDx-DR developed by Digital Diagnostics. IDx-DR uses AI technology to detect diabetic retinopathy, including macular edema.
- The Advanced Imaging and Modeling (AIM) lab uses technology to translate medical images into 3D and 4D interactive models to assist in pre-surgical planning of complex cardiac and cancer cases.
- The Children’s Innovation Lab, led by Adam Cross, MD, is prototyping a Pulmonary Acoustic Sensor Telemetry Array (PASTA). The device will have wearable adhesive pulmonary acoustic sensors to constantly monitor lung function while also classifying sounds that can indicate respiratory changes. Dr. Cross and his team must train the machine learning model with previously recorded lung audio from thousands of patients.
OSF Innovation’s work in the augmented intelligence space has received national attention.
Becker’s Hospital Review named OSF as a hospital and health system innovation program to know in 2023.
Also in 2023, AVIA, the nation’s leading market intelligence and advisory services firm advancing digital transformation in health care, has added OSF to its national Generative AI Strategic Collaborative.
In 2024, OSF joined 38 other health care payers and providers in signing the Healthcare AI Commitments, a collaborative effort with the White House to determine how to leverage frontier AI models to drive change in health care.
The future of AI
According to Dr. Handler, the use of AI is growing and changing so rapidly that it is hard to predict its future impact on health care trends. He doesn’t doubt, however, that it will be transformative.
In his words, “I hope that clinical decision-making, clinical workflows and access to care significantly improve over the next 10 years with the help of artificial intelligence. Will that actually happen? Only time will tell.”