In just 4 months, Moffitt Cancer Center’s new AI voice-based care team assistant, Mia (Moffitt Information Assistant), is surpassing the center’s expectations in streamlining the patient screening process before imaging appointments, improving patient satisfaction and reducing staff workload. Launched in February 2026, to help prevent unexpected delays on the day of scheduled magnetic resonance imaging (MRI) or computed tomography (CT) tests, Mia calls patients several days before an imaging procedure to ask a series of questions and review details about the test.

Unlike a prerecorded message, this system uses advanced, interactive voice AI-supported technology specifically for health-care communication that sounds surprisingly human-like when asking patients questions related to their specific imaging exam. For example:

• Do you have an implantable device such as a pacemaker or a nerve stimulating device?

• Are you allergic to contrast dye?

• Some patients undergoing an MRI screening may experience claustrophobia. Do you have concerns about being in an enclosed space?

The goal of the questions is two-fold: to prevent unexpected delays on the day of the scan and to ensure that the clinical team has all the necessary information on the patient’s health to perform the procedure. Patients can choose to opt out of talking with Mia, but refusal may cause delays in the procedure at the cancer center.

“Mia is designed to support both our patients and our care teams,” said Marc Perkins-Carrillo, MSN, RN, Chief Nursing Informatics Officer at Moffitt Cancer Center in Tampa, Florida. “We identified a pattern of last-minute imaging cancellations and delays driven largely by incomplete preprocedural screening and limited staff capacity to conduct outbound patient outreach. To address this, we implemented Mia as a scalable, automated solution to proactively collect critical screening information in advance of scheduled studies. This approach reduces avoidable cancellations and delays, improves operational efficiency, and enables our clinical teams to focus their time and expertise on patients who require more complex, high-touch care.”

Responding With Empathy

At the start of an imaging screening call with a patient, Mia introduces herself as an AI care team information assistant from Moffitt Cancer Center, specifies that the conversation is being recorded, and asks, “Is this a good time to go through a few quick safety questions?” All calls operate within secure HIPAA-compliant systems, and the information obtained is documented according to the cancer center’s clinical data standards and is securely maintained.

After the initial introduction, Mia proceeds to give details about the test and what patients need to avoid to complete the procedure. The information and questions are personalized for each patient based on the type of imaging test they have scheduled. For example, Mia explains the need to remove all jewelry and body piercings if the patient is undergoing an MRI scan; and informs patients undergoing a CT scan about the use of iodinated contrast dye to highlight any abnormalities and tumors, and asks if the patient has ever had a reaction to contrast dye in the course of a prior imaging procedure.

During a demonstration of Mia speaking with a patient scheduled for a CT test who had experienced nausea and vomiting, a common side effect of contrast dye, while undergoing a previous CT scan, Mia responded with what sounded like genuine concern and empathy. “Vomiting can be really uncomfortable,” said Mia. “I’m sorry you had to go through that experience. The care team will reach out to you to further discuss your contrast reaction.”

“Mia will interact with patients in a way that humans react in a similar situation,” explained Mr. Perkins-Carrillo.

Using AI to Improve Patient Care

According to Mr. Perkins-Carrillo, over a 7-week period, Mia completed calls with more than 10,000 patients, with 9,000 rating their interaction with Mia an 8.82 out of a score of 10. The bot identified approximately 22 patients with implantable devices who had to then schedule an appointment with their clinician to determine how best to proceed with an MRI test.

“Often when patients arrive for their imaging scan, they are unprepared for what the test entails and are sometimes unable to complete the procedure,” said Mr. Perkins-Carrillo. “For example, some patients might be concerned that they will experience claustrophobia while undergoing an MRI scan. If we are aware of that possibility ahead of time, the patient’s clinician can prescribe antianxiety medication, so the patient can feel comfortable proceeding with the test. And if we know in advance that a patient is allergic to the contrast dye used in CT scans, we can alert the clinician about the issue, so an alternative plan can be implemented. How to help patients be better prepared for these procedures is one of the biggest problems we were hoping to solve with Mia. And, so far, using this technology has been a success for both our patients and our clinicians.”

Using AI to Assist, Not Replace Clinicians

The AI-generated information assistant system for prescreening patients prior to an imaging procedure has been so successful, Mia recently took on a new task of preadmission testing screening to assess patients’ overall health before a scheduled surgery. Future plans include training Mia to become more proactive in her screening technique by integrating the AI-supported technology into Moffitt’s electronic medical record system to enable Mia to have access to information in patients’ health records.

“Instead of Mia asking patients whether they have an allergy to a contrast dye before a CT procedure, we would like to have Mia say instead, ‘I see that you have an allergy to contrast dye. Is that correct?’ And then send an alert to the patient’s clinical care team to resolve the issue before the patient arrives for the test,” said Mr. Perkins-Carrillo.

Although Moffit is exploring additional clinical uses for Mia, Mr. Perkins-Carrillo stresses that the technology is not a replacement for human care teams.

“We are looking at many more potential future clinical uses for Mia,” said Mr. Perkins-Carrillo. “Mia is a terrific support tool, but she is not a replacement for the cancer center’s medical care teams.”

DISCLOSURE: Mr. Perkins-Carrillo has no financial conflicts of interest to declare.

ASCO AI in Oncology is published by Conexiant under a license arrangement with the American Society of Clinical Oncology, Inc. (ASCO^®). The ideas and opinions expressed in ASCO AI in Oncology do not necessarily reflect those of Conexiant or ASCO. For more information, see Policies.