University Hospitals Excels at the Forefront of Artificial Intelligence and Healthcare, Enhancing Clinical Outcomes, Research, and Administrative Functions

UH Research & Educations Update | September 2024

Artificial intelligence (AI), the transformative data-driven technology that emulates human brain power, is revolutionizing the healthcare landscape, rapidly advancing diagnostic precision and treatment, and streamlining administrative tasks to enable more patient-centered care.

Propelled by immense high-quality data systems and analytics, AI empowers medical decision making, helping physicians and other care providers triage and treat patients in a timelier and effective manner, while maximizing opportunities to individualize treatment plans, amplify the precise cellular and diagnostic information available to care providers, and improve clinical outcomes.

“Big data inputs potentially allow us to be more accurate in identifying biomarkers for disease, radiologic and pathologic diagnoses, and personalized treatments,” says Daniel Simon, MD, President of Academic & External Affairs and Chief Scientific Officer and the Ernie and Patti Novak Distinguished Chair in Health Care Leadership for University Hospitals. “The volume of information is so vast that you need tools of big data, such as AI, for accurate analyses and definitive diagnostic and treatment recommendations.”

A Powerful Tool Reshaping the Healthcare Landscape

With AI-powered technology, physicians can better determine if a particular treatment will help a patient, or if the potential risks and side effects exceed the clinical benefit. For instance, at University Hospitals a combination of imaging for coronary artery calcium scoring and AI-guided methodologies is used to more precisely identify patients with a high cardiovascular disease risk who will most benefit from preventive therapies, such as statins and aspirin (i.e., 10-year risk of cardiovascular death, heart attack or stroke >7.5%). Classifying and treating patients according to quantified risk minimizes the potential for overtreatment based on guideline cut points.

AI platforms can also expedite the treatment of disease by specific genomic and cellular characteristics. For example, sequencing cancer patient genomic profiling with AI-assisted technologies readily hones in on mutations for which there might be a targeted or genetic therapy a patient might benefit from.

At a time when healthcare organizations across the country are challenged by workforce shortages, AI systems offer assistance and solutions for working faster, more efficiently, including built-in safety mechanisms that alert physicians and care providers to a change in a patient’s medical condition, says Dr. Simon. For example, machine learning algorithms and AI tools can track a patient’s blood pressure and heart and respiratory rates to warn medical staff when a patient is at risk of developing sepsis, prompting timely fluid resuscitation and antibiotic therapy to improve survival and limit complications.

AI Applications Multiply Amid Key Collaborations

“The heart of AI, what makes it so exciting, not only for healthcare, but for all other sectors, is the fact that it will automate many tasks that are repetitive or time consuming,” says Leonardo Kayat Bittencourt, MD, PhD, and Vice Chair of Innovation in the Department of Radiology. In medicine, AI will expedite a standard of care that is dependable, says Dr. Kayat Bittencourt, adding, “Almost everything we, human providers, do in healthcare can be helped and augmented by AI.”

Already, use of AI technologies has grown significantly throughout University Hospitals, across a range of clinical specialties and conditions. They are being used in the diagnosis and treatment of conditions such as pulmonary embolism, aortic dissection, vertebral compression fracture, and pneumothorax according to Dr. Kayat Bittencourt. Likewise, AI systems are helping physicians make treatment decisions in men with prostate cancer. They are also enabling radiologists to more quickly and accurately diagnose bone fractures in the emergency room and assist ophthalmologists striving to more precisely diagnose patients.

Strategic collaborations and partnerships with national and global industry have increased opportunities for research, innovation, and clinical applications across University Hospitals. In June, the health system announced the deployment of a comprehensive AI platform, Aidoc aiOS™, across the entire enterprise, including 13 hospitals and dozens of outpatient locations. This significantly advances AI system integration to enhance clinical outcomes and streamline workflow, providing access to 17 FDA-cleared AI algorithms.

In 2022, University Hospitals launched the Radiology AI and Diagnostic Innovation Collaborative, or RadiCLE, to advance the science of teaching, research discoveries, and clinical adoption of radiology AI, while also serving as a revenue stream for the radiology department and University Hospitals through collaboration with outside entities. This leverages the system’s wealth of data, advanced applications, diverse Northeast Ohio population, and world-class radiologists and specialists, to further advance the development and application of AI. The RadiCLE program is directed by Dr. Kayat Bittencourt, and the team includes a Program Manager, a Data Lead and a community of AI Clinical Champions composed of UH radiology faculty and a diverse representation of healthcare providers, trainees, and medical students.

Earlier this summer, the Department of Radiology at UH was designated an American College of Radiology® (ACR®) Recognized Center for Healthcare-AI (ARCH-AI). ARCH-AI is the first national artificial intelligence quality assurance program for radiology facilities nationwide. Becoming an ARCH-AI distinguishes University Hospitals as a leading-edge program in the adoption of best practices for the use of artificial intelligence technologies in radiology, affirming its position at the forefront of AI in

“The UH Department of Radiology is fortunate to have a team of world-renowned radiologists and informatics experts who are highly engaged in with the adoption and early-stage (AI) technologies, and they help drive us forward in their own areas of expertise,” says Donna Plecha, MD, Chair, Department of Radiology, Ida and Irwin Haber and Wei-Shen Chin, MD, Chair in Radiology, Co-Director, Diagnostic Institute, University Hospitals Cleveland Medical Center and Professor of Radiology, Professor of Pathology, Case Western Reserve University School of Medicine.

In addition to clinical applications, AI systems within University Hospitals are being used to magnify research opportunities. In 2023, the health system unveiled a collaboration with Premier Inc.’s PINC AI™ Applied Sciences (PAS) to advance clinical research, real-world evidence generation, and explore the development of new solutions, interventions and other novel healthcare initiatives using data-fueled AI platforms.

“Leveraging our combined data and research expertise will enable reduced research timelines, clinical care pathway development and testing, early disease intervention and help drive innovations in clinical, quality and operational effectiveness,” said Dr. Peter Pronovost, Chief Clinical Transformation Officer and Chief Quality Officer, University Hospitals.

Responsible Use of AI will Enable Continued Growth

Those integrally involved in the proliferation of AI throughout University Hospitals are quick to note the technology’s powerful capabilities, but equally mindful that the most powerful tool in medicine is an empathetic human touch. As use of AI in healthcare grows, physicians will remain the drivers of personalized, patient-centered care. Rather than replace physicians, AI will simply augment their capacity to treat patients, with vastly informed clinical judgement, in the most responsive, safest, and effective manner, says Dr. Simon.

“This is AI-assisted, AI-facilitated” care, Dr. Simon explains. “AI is a powerful tool that is enabling caregivers and health systems to be more responsive, accurate, effective and timely.”

To ensure patient privacy ample precautions are taken to remove patient-specific identifiers from the data used to develop AI systems, according to Dr. Simon. Strict data oversight at University Hospitals and HIPAA (Health Insurance Portability and Accountability Act) regulations require that any data provided to construct large data sets for AI systems be de-identified.

Moving forward, AI healthcare applications will undoubtedly proliferate, with greater capabilities and integration of comprehensive patient data to fully reflect heterogeneous patient populations, something University Hospitals is well positioned to do, given the diversity of patients it serves in Cleveland.

When it comes to future AI growth and identifying meaningful healthcare applications, “It is now widely accepted that AI is unlikely to fully replace us humans. However, humans that ethically embrace AI, and allow themselves to be augmented and be better for our patients, will certainly replace those who refuse to do so,” says Dr. Kayat Bittencourt.