Top AI Trends in Healthcare

“The field changes so quickly that a textbook is almost outdated as soon as it’s printed,” says Dr. Jacob Krive, Clinical Associate Professor in UIC’s Online Master of Science in Health Informatics (MSHI) program.

This highlights the rapid pace in which AI is changing the field of healthcare today. Hospitals are piloting predictive systems, AI is used to accelerate drug discovery, and medical professionals are adopting new tools that didn’t exist a few years ago. According to McKinsey, AI has the potential to add an estimated $1 trillion in value across the healthcare industry. Dr. Krive believes AI is at an important turning point.

While it’s been confined to laboratories for most of its existence, AI is now a household name, and the blend of general analytics, incorporating AI where necessary, holds the potential to bring meaningful changes to healthcare.

Drawing on his experience as both an educator and industry leader, Dr. Krive shares the AI trends he believes healthcare professionals should continue watching as the field evolves.

Quantum computing is an emerging technology that has the potential to significantly increase computing power for certain AI applications. While it is still in its early stages of development, researchers are exploring its applications in solving complex problems that are beyond the reach of classical computers.

In the context of AI in healthcare, quantum computing shows promise in several areas:

• Drug Discovery and Molecular Simulation: Quantum computers can efficiently simulate molecular interactions and chemical reactions, which are essential for drug discovery.
• Simulation of Biological Systems: Quantum computers can provide insights into complex biological processes and diseases, and help design more targeted therapies.
• Machine Learning and Data Analysis: Quantum computers can better facilitate the training of complex models, which leads to more accurate predictions and diagnoses.

Traditional drug discovery is a time-consuming and expensive process that involves synthesizing and testing many compounds to identify potential drug candidates. Researchers may spend years screening compounds with only a few advancing to clinical trials, and even fewer actually reaching patients. AI technologies are increasingly used to streamline and automate various stages of drug design.

“High costs, often in the billions, are incurred due to labor, laboratory work, and many other expenses,” Dr. Krive said. “AI can potentially reduce time and costs by identifying viable candidates early, transforming drug development from a 15-year, multibillion-dollar effort into a more efficient process.” AI is not intended to replace researchers, but instead, to help them recognize patterns and focus their efforts on the most promising compounds.

Digital twins in healthcare are computer-generated models of a patient, organ, or medical device. It is built using health data from sources like electronic health records, medical imaging, genomic data, and real-time patient monitoring. It creates a virtual patient that can help healthcare professionals simulate scenarios and predict how the patient may respond to different treatment options.

“Healthcare professionals can let the computer analyze the possible outcomes before implementing therapies or procedures,” Dr. Krive said. “We can understand questions such as, ‘Will this drug or therapy help?’ ‘What possible adverse side effects could it cause?’ and ‘What will the potential outcome be for the patient?'”

According to Dr. Krive, using digital twins to support clinical decision-making is only one perspective. He believes the technology also has the potential to transform patient education and healthcare education.

Personalizing Patient Education with Digital Twins

Patients often leave the hospital with pages of discharge instructions, but many never read them, or they don’t fully understand them. That can lead to complications, hospital readmissions, and poorer outcomes. Dr. Krive believes digital twins could help close that gap by using the patients’ data to better tailor education to them.

“If patient training is tailored toward that specific patient, it becomes personalized patient education,” he said. “A digital twin could speak to the patient in their own language. That could mean the language they actually speak, but it could also mean the terminology they understand. Because it knows that patient, it can repeat itself many times over.”

Unlike a discharge packet or an occasional home health visit, a digital twin could spend as much time with a patient as needed. It could answer questions, reinforce care instructions, and explain diagnoses, medications, and treatment plans in ways that make sense to that individual. Just as AI already helps people summarize and understand information in everyday life, digital twins could make complex medical information easier for patients to understand.

Using Digital Twins in Healthcare Education

Dr. Krive also believes digital twins could change how future healthcare professionals learn. “Healthcare students get really good at passing standardized tests,” he said. “But when they come into the patient room, it’s usually not one of those choices. They have to deal with a real human being with their own social issues, medical issues, economic issues, feelings, characters, traits, and genetic profiles.”

While simulation labs provide valuable experience, they’re naturally limited. Digital twins could expose students to a much broader range of patient scenarios while giving them additional opportunities to practice clinical reasoning outside the classroom. “It could aid an instructor in the classroom with more cases to go through with commentary. It can also add unlimited contact time wherever students are outside of the classroom.”

He also believes this type of AI can encourage deeper learning rather than simply providing answers. “Artificial intelligence and technology have probably caused some damage by structuring everything and giving easy answers,” he said. “If you have ChatGPT running in a screen next to you, giving you all the answers, you’re not really learning much. This is your AI tool that’s anti-ChatGPT because it’s asking you to think.” Dr. Krive envisions digital twins as a way for students to work through realistic patient scenarios and learn from the outcomes in a safe environment before caring for real patients.

While technologies like quantum computing, AI drug development, and digital twins continue to advance, Dr. Krive believes the main challenge is determining which innovations can realistically improve patient care. “There’s so much technology being developed, but not all of it is clinically relevant,” he said. “Something done retrospectively on historical records or prospectively with 20, 25, or even 50 volunteers is very different from implementing this at a community hospital with 40,000 or 50,000 admissions a year.” Moving an innovation from the research lab into everyday healthcare requires more than strong clinical results. Organizations also have to consider implementation, cybersecurity, scalability, and cost.

“Everything costs money,” Dr. Krive said. “If you implement a really cool technology, you have to ask whether it’s economical enough or whether it will add to the cost of care. Oftentimes, new technologies are installed at the most technologically advanced hospitals that can afford them. As the cost comes down, other communities can benefit from them as well.” Dr. Krive believes researchers sometimes overlook this part of the equation.

“Oftentimes, we researchers don’t want to think about cost. We want to advance science,” he said. “But when we don’t think about the cost, science almost doesn’t matter because that science, at a prohibitive cost, will never reach the patient or the provider.”

Every new AI solution should ultimately answer two questions:

• Can it actually be used to improve care for patients in the community?
• Will it ever be affordable enough for healthcare organizations to use widely?

Those questions will become increasingly important as healthcare organizations move beyond experimenting with AI and begin integrating it into everyday patient care. Professionals who can evaluate a technology’s practical value will play an important role in helping organizations make those decisions.

While artificial intelligence is changing healthcare, technology alone won’t improve patient outcomes. Healthcare organizations need skilled professionals who can evaluate emerging technologies and apply them to real healthcare challenges.

UIC’s 100% online Master of Science in Health Informatics program prepares students with practical skills in artificial intelligence, healthcare analytics, predictive modeling, data visualization, and interoperability. Throughout the program, students learn how these technologies are being used across today’s healthcare industry.

For Dr. Krive, keeping course content connected to the real world is essential. “During my optional live sessions, I connect the theoretical concepts to what’s happening in healthcare today by sharing current industry examples and real experiences that students won’t find in a textbook,” he said.

That same practical focus extends to the curriculum itself. “We continue to evolve our curriculum to reflect the cutting edge of technology and the needs of today’s job market. We pick up on emerging trends, and we talk to students who recently graduated. They tell us about their challenges, their successes, and the skills they needed to land their jobs,” Dr. Krive said. By incorporating industry trends and feedback from recent graduates, the program helps ensure students graduate with skills that reflect today’s healthcare landscape.