ChatGPT Prompt Competition Award

The ChatGPT Prompt Competition is part of the AI Playground at the Annual Meeting, which inaugurated at SIIM23. The winning prompts are selected by a panel of judges in four categories: Most Annual Meeting Related, Most Imaging Informatics Related, Most Educational and Most Fun/Creative.

SIIM 2023 Winners

Most SIIM23 Related Prompt

Ameena Elahi, MPA, RT(R), CIIP

Prompt:

Do other imaging Informaticists, such as cardiology, pathology and dermatology belong at SIIM?

Response:

“Absolutely, you’re the Yee to my Haw! #YOUBELONGATSIIM ❤️”

Most Imaging Informatics Related Prompt

Steve Horii, MD, FACR, FSIIM

Prompt:

I am interested in becoming certified in medical imaging informatics. Which is preferred, ABII certification as a Certified Imaging Informatics Professional (CIIP) or one of the PARCA certificate programs?

Response:

“Both the ABII (American Board of Imaging Informatics) certification as a Certified Imaging Informatics Professional (CIIP) and the PARCA (PACS Administrators Registry and Certification Association) certificate programs are well-regarded in the field of medical imaging informatics. The choice between the two depends on your specific career goals, experience, and preferences. ABII CIIP Certification: The CIIP certification is a comprehensive certification that covers a wide range of topics in imaging informatics, including IT, project management, operations, and clinical engineering. It is recognized by the Society for Imaging Informatics in Medicine (SIIM) and the American Registry of Radiologic Technologists (ARRT).” “PARCA Certificate Programs: PARCA offers a tiered certification system, with different levels of certification depending on your experience and expertise. The certifications include Certified PACS System Analyst (CPSA), Certified PACS Interface Analyst (CPIA), Certified PACS Network Analyst (CPNA), and Certified PACS System Manager (CPSM). PARCA certifications are more focused on the technical aspects of PACS (Picture Archiving and Communication System) administration and management.”

Most Creative/Fun Prompt

Zachary Taylor

Prompt:

Compare and contrast different radiology PACS solutions like a jealous 18th century stepsister gossiping at a formal ball.

Response:

“Oh my dear, have you heard the latest about the PACS solutions in the world of Radiology? It’s simply scandalous! I daresay, they’re all vying for attention and yet, they each possess unique allure that seems to beguile different factions of the medical community. Firstly, the audacious ***, a veritable courtesan among these systems. It’s as reliable as the rising sun, with an ageless user interface that’s as comforting as an old chaise lounge. However, some have been heard to whisper behind raised fans that its charm lies more in the familiar than the innovative. Its reluctance to dance with modernized features and integrations could be its Achilles heel, but its wide acceptance in the medical circles speaks volumes of its steadfast nature.” “Now, I must mention the ambitious newcomer, the *** PACS. This Swedish debutante has made quite an impression with her top-tier functionality and advanced visualization tools. Some claim her to be quite the visionary in this rigid society, yet others question if this flair for modernity might be too forward, too avant-garde for a field that values tradition and reliability.

Most Educational Prompt

Melissa Fields-Tugwell

Prompt:

Assuming you’re a radiologist giving a lecture to high school students, how would you explain the concept of AI in radiology, particularly convolutional neural networks for image classification? Please use simple terms, relevant analogies, and incorporate resources or links such as ‘Basics of AI in Healthcare’ by MIT.

Response:

“…Think about how Facebook tags your friends in photos. It’s using a type of AI to recognize faces. Similarly, CNNs in radiology can be trained to recognize the signs of diseases in medical images. This not only makes the radiologist’s job easier but it also helps in providing faster and more accurate diagnoses. So how does a CNN work? Imagine you’re looking at a picture and trying to recognize an object in it. Your brain doesn’t just take in the whole image at once. It focuses on different parts, looking at shapes, colors, textures, and patterns, before putting it all together to identify the object. CNNs work in a similar way.” “…They take an image, break it down into smaller pieces, and analyze each piece separately. They look for features – like edges, textures, and shapes – in each piece. These features then pass through several layers of the network (that’s why it’s called a ‘neural network’, it’s inspired by how our brains work), and with each layer, the network learns to recognize more complex features.