AI at UC Davis Health

Today, artificial intelligence (AI) is already at work improving outcomes in health care – from increasing efficiency to improving diagnoses to accelerating the discovery of more effective treatments, AI is transforming the way that healthcare is delivered.

At UC Davis Health, AI makes recommendations, and your doctors make the final call.

We’re using AI to provide safer, faster and more personalized care for our patients. With the help of AI, we can reach even more patients, expand access to care, and create new ways to diagnose, treat, prevent, and cure disease. We’re delivering advanced care, informed by AI, in the hands of local expert physicians.

Keeping Patients and Providers at the Center of What We Do

UC Davis Health has always focused on providing the best care possible to our patients and community, using the latest in medical techniques and technology. Now, AI enables a whole new set of tools to improve the health of our patients, reduce burdensome work for our care providers, make the delivery care more convenient and patient-focused, and improve communication between providers and patients. We view AI as another doctor-supervised technology, and one of many technology tools that allows us to deliver advanced care, guided by our local experts.

We know the concept of ‘AI-assisted care’ can cause concern for some people – and, at the same time, excitement, and optimism for other people. Our philosophy is that AI should assist in patient care, improving what our providers can do for patients, but it should not replace the ultimate decision-making of providers and patients themselves. We are committed to using AI responsibly, ensuring that AI preserves the patient-provider relationship, and improves the effectiveness of care delivery and outcomes, while staying true to our values of kindness, equity, and compassion, so we can deliver tomorrow's care today.

UC Davis Health is at the leading edge of AI, both in terms of the AI we adopt and how we test and track results before, and after, we adopt it. We have very clear goals for the use of AI in healthcare, including improving outcomes (by reducing inaccuracies and identifying issues earlier), improving the physician-patient relationship (by reducing screen time during visits), better personalizing health services, and automating behind-the-scenes tasks (such as coding and summarizing notes).

The bottom line for us is that, while AI may assist us in how we deliver care, doctors and their patients make the final call about medical decisions. AI is helping our doctors and nurses deliver tomorrow’s care to our patients today.

Some Examples of Innovative AI That We Have Adopted

Imaging – Viz.ai; Lunit mammography
Augmented Reality for spine surgery
Generative AI/NLP - Illuminate; TIGER Aortic Stenosis detection
Queen of Hearts STEMI detection MD over-read
Custom Pop Health predictive models – ED/Hosp admissions, need for emergent HD

Examples of How We Make Sure That We Adopt AI Responsibly

AI Oversight Committee

Made up of health professionals from all fields, this committee uses rigorous methods to assess the safety, accuracy, and fairness of the AI we use at UC Davis Health. The committee is charged with ensuring that AI is used responsibly and ethically to serve the needs of our patients, providers, students, and staff.

Learn more about the AI Oversight Committee

AI Health Equity “Task Force”

Led by Dr. Hendry Ton & Dr. Reshma Gupta, this group is a group of health equity experts, clinicians, statisticians, and data scientists that develop the standards used to make sure that AI is developed and adopted equitably, ensuring that AI meets the needs of all members of our community.

BE-FAIR Framework

BE-FAIR stands for Bias-reduction and Equity Framework for Assessing, Implementing, and Redesigning (BE-FAIR) healthcare predictive models. BE-FAIR provides guiding principles that promote anti-racism, community engagement, inclusion of historical context, and equity-focused AI development best practices when AI is developed and implemented at UC Davis Health. It’s one of the many ways we work to ensure that AI is adopted responsibly in service of our patients and community.

Learn more about BE-FAIR

Health Analytics Core Team

This inter-disciplinary team of healthcare providers and data scientists analyze potential AI applications and their use prior to adoption to evaluate whether they are safe enough for patients at UC Davis Health. After a trial period, the group reviews performance to ensure outcomes match expectations. UC Davis Health also develops our own custom AI apps, where doctors, nurses, and other health professionals work together with the Health Analytics Core to develop solutions that are custom designed to address challenges specific to our patients.

Learn more about the Health Analytics Core team

State of the Art Security

We have a data security team that reviews, and pressure tests, AI applications to ensure that AI adoptions prioritize patient privacy and keep patient data safe.

Data Sharing Committee

This multi-disciplinary committee of clinicians, scientists, privacy and security experts, and legal specialists sets policies and guidelines to ensure data are not shared with third parties, unless the appropriate approvals and safeguards are in place, and ensures that only approved AI products are used at UC Davis Health.

Learn more about the Data Sharing Committee

Patients at the Center of AI

AI is helping doctors and nurses focus more on what matters most – caring for patients and delivering faster, more convenient, and more comprehensive care. The use of AI at UC Davis Health is designed to support and enhance the work of doctors, nurses and other providers – not replace them. Here are some examples of how we’re putting AI to work for our patients:

Faster, More Accurate Diagnoses

AI is Improving Patient Outcomes

Our AI tools are aiding in stroke care, sepsis detection, and surgical decision-making – leading to faster interventions and better survival rates.

Medical Imaging

AI-powered software helps our radiologists quickly and accurately analyze complex images, such as X-rays and MRIs, flagging potential issues for rapid expert review. Even with these reviews, human experts are still independently reviewing all images and making the final decisions about actions to be taken.

Predictive Analytics

Our teams use AI to analyze large datasets to identify patients at high risk for certain conditions, like sepsis and emergency department visits, allowing for proactive intervention before a crisis occurs.

Enhancing Patient Care

Reducing Administrative Burden

AI-powered tools such as Abridge assist our clinicians with administrative tasks, such as creating visit summaries and managing patient data. This frees up more time for doctors to focus on what matters most – you.

Optimizing Hospital Operations

AI helps us streamline hospital logistics and scheduling to improve patient flow and ensure resources are available when and where they are needed.

Using AI Responsibly to Care for Patients

When it comes to AI in health care, issues like fairness, privacy, accountability, and transparency matter. UC Davis Health has established a rigorous framework to ensure all AI technologies are used responsibly and ethically, with doctors and nurses in charge, using tools that have been reviewed and verified by doctors, with humans making the final decisions in AI-assisted care.

1. A Human-Centered Approach

We have guidelines for using Generative Artificial Intelligence that protect patient privacy and data security while also prioritizing accuracy and checking for biases.
Our patient-first AI Oversight Committee includes broad organizational representation and the technical expertise required to evaluate the safety, accuracy, fairness, and appropriateness of proposed AI models. We have developed rigorous methods to evaluate if AI is fit for adoption and behaving as expected after it’s implemented.
We believe that AI is a tool to assist, not replace, human judgment. Every AI recommendation is reviewed and validated by our expert clinicians to ensure the highest standards of patient care and human dignity.

2. Data Privacy and Security

HIPAA Compliance: We adhere to the strictest privacy regulations, including HIPAA, to protect your health information.
Anonymization: For research AI development, we use advanced anonymization techniques to ensure patient data is de-identified so an individual cannot be re-identified through its use.

3. Addressing Bias and Promoting Equity

We actively work to identify and mitigate bias in our AI algorithms by using diverse datasets that are fully tested and vetted.
As AI programs are used, we evaluate their performance to make sure AI technologies are not contributing to disparities in outcomes for any of our patients.

AI Innovations Benefit Patients

Across UC Davis Health, our medical scientists are developing and testing new ways to use AI to care for patients. These tools can sort through immense amounts of data to create promising treatments and help predict how they might work for different patients. It’s local doctors using the latest AI tools for better patient care, and here are some examples:

Life-saving Aneurysm Care

In 2022, UC Davis Health used AI to scan 3 million patient records of patients with abdominal aortic aneurysms who had been ‘lost to follow up.’ AI identified 1,000 patients in that group that needed attention – 643 of those people had clinic visits for follow-up, 81 people ended up having surgery, and of those, 32 lives were saved through additional follow-ups. That wouldn’t have happened without AI.
Equity-minded Predictive Care

We created an AI-driven population health predictive model to identify patients at risk for potentially preventable emergency department visits and hospital admissions who may benefit from care services that prevent them from getting sick. We used our home-grown BE-FAIR AI equity framework to develop the predictive model responsibly, ensuring that health equity is embedded at every stage of AI development and implementation.
Read more
Improving Patient-Physician Relationships

Our clinicians are now using Abridge, a technology that allows them to focus more directly on patients and their health needs – and which has made doctors 12% more efficient so they can spend that time with their patients instead of their computers. Abridge records patients’ conversations with their clinicians during visits, then creates a summary using AI. After the visit, the clinician reviews the summary for accuracy and adds it to the patient’s medical record.
Learn more about Abridge
Anticipating Strokes

We are using a technology platform, Viz.ai, to help quickly identify patients suspected of having a stroke. UC Davis Health is the first in the Sacramento region to use the platform. It utilizes image-based AI to analyze a patient’s computerized tomography (CT) scan and alert care teams of a potential stroke within minutes. Our stroke specialists then review the CT scans and can initiate life-saving care faster than ever before.
Learn more about Viz.ai
More Accurate Mammograms

Every mammogram at UC Davis Health is read by both a radiologist and an AI detection algorithm. The algorithm has been shown to increase cancer detection and decrease false positives. Also, the health system is co-leading a $16 million national clinical trial to evaluate whether AI can help radiologists interpret screening mammograms more accurately. The goal is to improve breast cancer detection and reduce unnecessary callbacks and anxiety for patients.
Read more

Contact Us

For specific questions about your care, please contact your physician. For information about partnering with us on AI initiatives, please contact our Ventures team.

Donate to support our AI research

Partner with us in our AI research

Work with us on AI development

Information for potential partners and AI companies

Disclaimer: This page provides general information about our use of AI. It does not provide medical advice. Consult with a UC Davis Health care professional for diagnosis and treatment.