AI Certificate Program Builds Practical AI Expertise
Participants say training strengthens skills in AI evaluation and implementation
As AI becomes more deeply integrated into clinical radiology practice, radiologists and medical imaging professionals who completed courses from the RSNA Imaging AI Certificate Program say the training helped them better evaluate, implement and monitor AI tools.
For those who have completed the courses, benefits ranging from a stronger understanding of AI fundamentals to deeper knowledge of validation, deployment and performance assessment. Available certificate options include Foundational, Advanced, Emergency and Chest, with a Neuroradiology Certificate course scheduled for release in late 2026.
“The Foundational curriculum is an opportunity for learners eager to stay ahead of the AI curve,” said Robert Freed, MD, a radiologist specializing in abdominal and emergency Radiology at Midwest Radiology based in Minneapolis.
“Despite having some baseline AI knowledge, I appreciate how the modules in the Foundational curriculum helped me level-set my grasp of AI,” Dr. Freed said. “This includes AI in radiology but also as a broad overview of the basic premise of the algorithms: how they learn and grow, and their limitations.”
Dr. Freed explained that one concept emphasized in the Foundational curriculum is AI “drift,” which can occur when an algorithm is applied in circumstances different from those in which it was trained, potentially affecting performance.
To illustrate why radiologists need that understanding, he compared them to NASCAR drivers.
“Radiologists are NASCAR drivers and AI is the souped-up engine growling under the hood. We don’t necessarily need to know how to build the algorithm or engine, but we’re going to use it every single day,” Dr. Freed explained. “We need to know its tendencies and flaws, and when we get an output that feels ‘off,’ we must be able to communicate that and develop an intuition to recognize when there’s something wrong with the output."
“AI alone, or radiologists alone are imperfect, but when you combine AI with baseline content knowledge, the synergy is off the charts,” he emphasized.
Dr. Freed has completed three certificates in the program and said he believes AI could reshape expectations for radiologists and influence standards of care.
Advanced Course Supports AI Implementation
Polina Pilius, MD, project manager and lead technical radiologist for V-Star, a U.S.-based medical annotation company, said the Advanced course reinforced the importance of defining clear objectives when implementing AI.
“The instructors underscored how important it is for us to define the main idea or main question and what we want to answer using AI,” she said. “The training is useful for medical specialists no matter if they already work with AI or they just expect to work with it in the future.”
As the company’s AI work continues to expand, she said their leadership wanted to support physicians involved in those efforts. “That’s why this opportunity to learn from the course was very useful,” she said.
After taking the Advanced certificate course, Dr. Pilius said she learned a lot about implementation processes and how to validate and test the results of AI. “It is exactly what I’m doing in my project practice,” she said.
Learning about radiomics was particularly valuable to Dr. Pilius, who also works at the Research Institute of Oncology and Radiology in Kazakhstan. She was already familiar with radiomics in theory, but lacked an understanding of its practical application, as well as the details of conducting radiomic analysis in oncology.
Working through the topic on her own had been difficult, she said, whereas the course presented everything clearly and intuitively. Drawing on the knowledge gained from the course, Dr. Pilius gave lectures at the Institute for physicians participating in AI research and was able to prepare a methodological guide on AI in biomedical research to support her colleagues.
Building Skills Across Career Stages
The program also supports learners at different stages of their careers, from trainees building foundational expertise to practicing radiologists implementing AI tools.
Nikolaos Koutsogiannis, MD, MSc, a private practice radiologist in Athens, Greece, entered the program with a background in AI gleaned from his academic coursework and residency experience. While contemplating a thesis topic for his master’s in biomedical imaging and radiation protection, he came across the Radiology: Artificial Intelligence article, “Checklist for Artificial Intelligence in Medical Imaging (CLAIM),” which sparked his interest in AI applications in radiology.
Seeking a stronger grounding in AI, he enrolled in the Foundational course. “Observing the transition of a model from a research context into actual clinical practice was exceptionally enlightening,” he said. “It motivated me to use the CLAIM checklist for my master’s thesis to assess the reporting quality of deep learning studies.”
He has since completed four certificates, surpassing his initial intent. “As my academic and professional interests evolved, I found a curriculum that matched each stage of my journey,” Dr. Koutsogiannis said.
The Advanced course deepened his understanding of algorithm architecture, data bias, and data curation. With his residency institution specializing in emergency management and acute care, while also implementing AI-driven lung nodule evaluation, his progression to both the Emergency and Chest certificates was a natural next step.
“The initial exposure established strong familiarity with AI capabilities and clinical applications,” Dr. Koutsogiannis said. “Completing all four certificates successfully transitioned my knowledge into an advanced, structured competence. This training has empowered me to clinically evaluate vendors, spot algorithm errors and safely integrate these tools into daily practice.”
Drs. Pilius and Koutsogiannis note that the certificate program helps make the technology more approachable and practical. “Each certificate program delivers specialized expertise that can be seamlessly integrated into daily clinical workflows to improve diagnostic accuracy and optimize operational efficiency,” Dr. Koutsogiannis said.
Dr. Pilius believes that understanding will become increasingly important as AI adoption grows across radiology. “I don’t imagine a future without AI implemented into the radiology field,” she said. “That’s why understanding is the best tool to counteract hesitancy.”
“Rather than being a distant futuristic concept or an uninterpretable ‘black box’, AI should be approached through the lens of explainable AI,” Dr. Koutsogiannis emphasized. “This is crucial for clinicians, as it provides the necessary transparency to transform deep learning models into practical tools that address immediate clinical challenges and fundamentally prioritize patient care and safety.”
For More Information
Learn more and enroll in the RSNA Imaging AI Certificate Program.
Read previous RSNA News stories on AI education:
- Understanding the Capabilities and Limitations of LLMs in Radiology
- AAPM, ACR, RSNA and SIIM Announce Joint Effort to Develop Innovative AI Educational Framework for Radiology
The RSNA Imaging AI Certificate Program Update
The RSNA Imaging AI Foundational Certificate v2.0 helps imaging professionals build practical skills for evaluating, implementing and monitoring AI tools. The refreshed curriculum features new clinical use cases and practice-ready insights designed for real-world workflows, helping participants make informed decisions about AI adoption and use.
Developed by radiologists for radiologists, the program focuses on the knowledge needed to confidently apply AI in clinical practice.
Additional pathways in the program include Advanced, Emergency and Chest certificate courses with a Neuroradiology course planned for late 2026. Learn more and enroll now.