Getting Started With Artificial Intelligence

Learning about artificial intelligence (AI) can be an intimidating task, especially without access to mentors or resources to help navigate this complex domain. However, an increasing role of AI in the practice of radiology necessitates familiarity with this technology to ensure its effective and ethical use.

Recent literature identifies barriers to AI education and provides solutions to guide initiatives tasked with democratization of AI education. Widespread integration of AI topics in radiology training curricula will take time, requiring trainees with an interest in imaging AI to pursue alternative routes for obtaining AI education. The following list highlights suggestions for navigating this complex landscape.

Identify Your Goals

The scope of imaging AI is daunting, considering the growing, vast array of image interpretive and non-interpretive use cases. In this context, it is crucial to reflect on your goals for pursuing AI education to define specific steps that will guide your endeavors. For example, trainees desiring to create AI solutions may start with the basics of machine learning (ML) followed by advanced training in technical ML topics covering architecture, coding and emerging paradigms in ML technology.

Alternatively, trainees interested in facilitating clinical application of AI solutions or serving primarily as end users may start with the basics of ML, shifting focus to topics central to clinical integration, such as machine learning operations (MLOps), standards and interoperability, and practices for monitoring AI.

The roles of tool creators and users are not mutually exclusive; however, limited trainee time for AI education requires a strategy that prioritizes specific trainee interests within the realm of imaging AI.

Find a Mentor

Mentors provide vital support in the form of expertise and guidance for trainee AI endeavors. Availability of mentors for AI education will vary across training programs. Trainees without access to local mentors should participate in national and international initiatives that eliminate physical barriers between trainees and AI experts, such as the National Imaging Informatics Course – Radiology (NIIC-RAD) and RSNA Imaging AI Certificate Program.

Curate Resources

Curating resources for your imaging AI journey can be cumbersome, often requiring a trial-and-error approach to find those that best fit your level of expertise and AI education goals. Instead of pursuing a “shotgun” search for sources on the internet, start with this list of compiled introductory resources as a reliable starting point.

Network/Collaborate

The imaging AI community embraces collaboration and features approachable experts with a vested interest in trainee education. Attend virtual and live gatherings, such as conferences or online webinars, to meet and learn from members in this vibrant community.

Keep Up With Latest Content

Imaging AI is a rapidly evolving field, necessitating a proactive approach to stay informed about new paradigms and use cases. Several outlets, such as The Imaging Wire, address this need by providing routine updates about the latest research and industry news in concise formats with options for more advanced reading. Subscribe to these resources to avoid falling behind on developments in imaging AI.

An interest in AI is the first step to pursuing AI education and effectively using AI solutions in practice. Hopefully, the steps and resources outlined in this article will aid in getting you started on your journey.