On Artificial Intelligence: An interview with Gennaro D’Anna

The latest addition to our “On Artificial Intelligence” series finds us diving into artificial intelligence (AI) within the world of radiology and, more specifically, neuroradiology. Gennaro D’Anna, a neuroradiologist at CDI Centro Diagnostico Italiano in Milan, Italy, discusses challenges to adoption in clinical practice, what the future looks like for AI, and what new generations of radiologists can do to familiarize themselves with AI tools. Let’s dive in! What is your background/experience with artificial intelligence and what first attracted you to the topic? I’m an experienced Neuroradiologist involved in hospitals and outpatient diagnostic centers. I have extensive experience integrating artificial intelligence into clinical practice. Over the years, I have evaluated a diverse range of AI software in real-world settings and have also helped in the development of AI algorithms tailored for neuroradiological applications. I remain actively engaged in the rigorous assessment and validation of AI tools to enhance diagnostic accuracy and patient care in neuroradiology. Before embarking on my career as a neuroradiologist, I was a passionate admirer of science fiction and an avid gamer. This early fascination with futuristic worlds, where machines can help mankind in several jobs, made the prospect of working with artificial intelligence feel like a natural progression. As a clinical neuroradiologist, I quickly recognized the transformative potential of AI in helping with massive radiological and neuroradiological data. The speed that AI can offer in evaluating abnormalities inside the huge number of images now produced with the new technologies will benefit our field. Moreover, the integration of sophisticated algorithms into clinical practice—enhancing workflow and improving patient outcomes—aligns perfectly with my longstanding interest in innovative, technology-driven solutions. This convergence of personal passion and professional responsibility has been a key driver in my commitment to advancing AI in neuroradiology. Finally, commitment of societies like EuSoMII on highlighting the importance of AI also contributed to my “attraction” to the topic. What are the biggest challenges to AI adoption in clinical practice? One of the most significant challenges to AI adoption in clinical practice is infrastructure. Many healthcare institutions are hampered by outdated IT systems that require substantial upgrades to support modern AI applications effectively. In addition to these infrastructural issues, education is equally critical. There is considerable misinformation and varied perceptions about AI’s capabilities and limitations, as well as misunderstandings regarding cybersecurity, data sharing, and safety. These discrepancies lead to different levels of understanding and acceptance, causing communication barriers within teams. To achieve meaningful large-scale adoption, it is essential to have well-trained and knowledgeable staff as well as appropriate resources to support the integration and maintenance of AI tools. Give us an example (or an educated guess) of what you think AI will be able to do in 3 years. What about in 10 years? I have to explain this more as a premise: I’m not an oracle! In the next 3 years, I think that every radiology machine will be equipped with advanced AI capabilities, seamlessly integrating into clinical workflows to enhance both image acquisition and interpretation. Over the subsequent decade, AI is expected to assume a substantial role in routine screening processes, significantly improving efficiency and productivity by handling a large share of initial evaluations. Of course, these are merely possible scenarios, and many other groundbreaking applications and discoveries remain on the horizon. Moreover, greater technological advancement necessitates effective governance; as physicians, we must ensure that technology not only augments our diagnostic capabilities but also enriches our relationship with patients, ultimately leading to a more compassionate and globally improved healthcare system. As a radiologist, what is your advice for younger colleagues wanting to dive into the topic of Artificial Intelligence, Machine/Deep Learning, and/or Radiomics? What are your tricks for staying up to date in this fast-evolving field? I fully agree with Brendan Kelly’s previous answer: a radiologist, whether young or experienced, doesn’t need deep knowledge of all the hidden mechanisms inside an AI “black box”. However, it is essential to stay open and curious about the field, at least by having a kind of “Hitchhiker’s Guide for AI Enthusiasts”—a structured curriculum that can be integrated into residency training and supplemented by courses from esteemed scientific societies. With this foundation, understanding the most practical aspects of AI becomes much easier. Brendan gave a great example with MRI: I love spending time in the scanner area with technicians, trying to get the best out of the machine, but I don’t fully understand all the physics behind it (and, unfortunately, I’m not a coder). Of course, those who want to dive deeper into technical aspects of AI are more than welcome to do so—they will become an essential link between radiologists and industry. For everyone, staying up to date is even more crucial. Keeping up with the most important journals is not just recommended—it’s mandatory.