Artificial intelligence, or AI, is bringing innovation to many industries, including health care. AI technology and tools process data, recognize patterns and can make decisions or provide recommendations based on the information. In 2025, the Food and Drug Administration authorized more than 1,000 AI-enabled devices. According to a survey of doctors by the American Medical Association, two thirds of doctors are choosing to use AI in their practice, and that number is expected to grow. When AI is used thoughtfully and responsibly, it has the power to enhance a patient’s care and their experience.
“In many cases, we are adopting AI technology to help provide the highest quality care possible, from accurate, timely diagnoses to the experience of receiving care from our clinicians and teams,” said Brooke Crotty, MD, MPH, an internal medicine physician and interim president for Inception Health, the innovation arm of Froedtert ThedaCare.
As part of Froedtert ThedaCare, the Froedtert & the Medical College of Wisconsin health network and ThedaCare strive to be your partner in living your unique, best life by transforming health care through excellence and discovery. While embracing what AI has to offer, Froedtert ThedaCare has procedures and controls in place for using AI for patient care to keep patients safe.
Here are just a few examples of how our patients are already benefitting from AI in their care:
AI-Assisted Medical Imaging
In health care, one of the most promising and widespread uses of AI is in radiology. Using AI to assist clinicians with identifying abnormalities in diagnostic imaging can lead to faster results that may reduce patient anxiety and lead to quicker treatment.
Emergency Treatment
In many cases, care teams can use AI tools as they analyze X-rays, CT scans and MRIs. The technology can shorten the time to a diagnosis as well as alert clinicians to subtle red flags in scans that the human eye might not instantly see. For example, in the case of a brain bleed found on a CT scan, AI can alert a radiologist to review the image ASAP, which could reduce time to surgery or other treatment. In critical care settings, like an emergency department, this helps keep patients moving to the care they need more quickly and frees up beds for incoming patients.
Breast Cancer Risk
Care teams also use AI to help assess breast cancer risk. Mammogram images can be run through an AI tool that provides a risk score for the radiologist to review. The score is used to help identify women with a higher risk of developing breast cancer. The tool flags these patients because it was trained on hundreds of thousands of mammogram images to look for the specific patterns that suggest higher risk.
“These patients can be referred to our high-risk breast cancer clinic for additional care that may include monitoring, genetic testing, surgical options, medication or lifestyle changes,” said Debbie Cray, vice president and chief data and analytics officer for Froedtert ThedaCare. “Using AI helps us elevate our prevention efforts against the disease.”
Several studies have shown that AI is more accurate in predicting breast cancer risk than other breast cancer risk calculators.
AI for Clinical Trial Matching
Clinical trials give patients access to treatments that may not be available elsewhere, but matching people to the right trials can be a challenge. It takes time, is complicated and relies on people to do the work. Led by Anai Kothari, MD, a Froedtert & MCW surgical oncologist, doctors at Froedtert Hospital are using a tool called OncoLLM that matches patients and trials with the goal of giving patients more opportunity to quickly take advantage of potential treatment options. OncoLLM securely analyzes all available clinical data, including clinician notes in the medical record, lab results and imaging, to assess each patient’s eligibility for open clinical trials.
“Clinical trials are a key part of optimizing cancer care but only when patients have opportunities to participate in them,” Dr. Kothari said. “Using an AI-powered model like OncoLLM helps us make sure that patients are being assessed for all the possible trials we have open. This could transform the clinical research process and ultimately, improve overall outcomes.”
OncoLLM is being used for gastrointestinal (GI), breast, thoracic (lung), prostate and other urologic cancer patients. Early results of the pilot are promising, and eventually, OncoLLM may be used for clinical trial matching across all types of cancers in the health system.
Ambient AI for Medical Documentation
Ambient-listening and transcription technology powered by AI can simplify how information from a doctor visit goes into a patient’s electronic medical record. AI is used to save time by capturing the conversation during the visit and converting it into notes, after-visit summaries, referrals and prescriptions. Clinicians can use their cell phone as a portable microphone to capture the conversation and document the information for the medical record.
“When really good technology works, it fades into the background,” said Dr. Crotty. “While the ambient scribe takes care of typing up the notes during the visit, the doctor has more time, face to face, for the actual conversation with their patient.”
The ambient scribe notes are available at the end of the visit for the clinician to review and confirm accuracy.
Using AI Responsibly
Froedtert ThedaCare takes many steps to use AI responsibly. Protecting patient privacy is a top consideration through targeted oversight, regulatory and ethical compliance as well as ongoing risk assessment.
“We have the governance and controls in place to monitor the data flow, and we work hard to make sure AI is used in a safe, secure, HIPAA-compliant way,” Cray said. “Our goal is that our data is only used to benefit our health system and our patients.”
Froedtert ThedaCare has an AI governance board of experts who evaluate use cases for the technology, make sure it is thoughtfully made part of the health network’s operations and monitor its use over time.
The health system’s use of AI is guided by the following principles:
- Human judgement is essential. AI should support health care professionals, not replace their clinical or operational judgement
- AI tools should be designed and used to protect patient privacy and data security
- AI can have biases and needs to be regularly monitored and tested to make sure it works fairly for everyone
- Collaboration on how AI is used is a must. Decisions about it should include care teams, ethicists and technology and data experts with an understanding of AI
“AI takes away some of the drudgery in clinical practice, freeing up more time for our clinicians to be clinicians,” Dr. Crotty said. “The computer is not replacing the doctor. It is a tool to assist, so doctors can offer the most support and most personalized care to patients.”
For more information about innovation in health care, visit froedtert.com/innovation.
