Artificial Intelligence for ophthalmic research

(included with All-Access Saturday pass)

Saturday, May 3
8am – 5pm MT
Calvin L. Rampton Salt Palace Convention Center

Organizers

Michelle Hribar, MS, PhD; Aaron Lee, MD, MSCI; Jayashree Kalpathy-Cramer, PhD and SriniVas R. Sadda, MD, FARVO

All the things you wanted to know about AI but were afraid to ask

8 - 10am
View the agenda

Target audience
Beginner-level: No prior experience with AI, programming or tools is needed.

Description
For those who want to learn about artificial intelligence (AI) but are not interested in learning how to create it. The basics of AI in plain language will be introduced. Didactic lectures will explain AI, its fundamental concepts and related terminology. The last hour will be devoted to an open Q&A with the audience. Come prepared to get the answers you want and need. Already have a question in mind? Want to ask a question anonymously? Send it to the organizers now via our question collection form.

Large language models and how they can be used for research – Hands-on lab

10:30am - Noon
View the agenda

Target audience
Intermediate level: Attendees should be familiar with the basics of AI and be seeking to use existing tools in their clinics or research. 

Description
Will introduce large language models, review their strengths and limitations, and provide hands-on instruction for how to use them in your research projects. Participants need to bring their own laptop for the hands-on portion of the session. 

Equipment requirements
A Google account, laptop and headphones are required to participate in the hands-on exercises. Related files will be sent via email at least three business days prior to May 3. 

Fundamentals of model training and federated learning

1 - 3pm
View the agenda

Target audience
Advanced level: Attendees should be seeking to create AI models themselves and be familiar with basic programming skills.

Description
Will introduce the basics of programming models as well as how to build an AI ready dataset. Attendees will learn also the basics of federated learning — an approach to building AI models from multisite data where the data can stay local to each site.

Training a deep learning model – Hands-on lab

3:30 – 5pm
View the agenda

Target audience
Advanced level: Attendees should be seeking to create AI models themselves and be familiar with basic programming skills.

Description 
This session will provide hands-on lab exercises focused on training deep learning models, emphasizing practical implementation and analysis of model performance. Participants will actively train models, monitor training curves, and learn techniques for diagnosing common training issues such as overfitting and underfitting. Additionally, we will introduce concepts related to federated learning, highlighting how models can be collaboratively trained across decentralized data. By the end, attendees will have gained practical skills in training deep learning models and a foundational understanding of federated learning.

Equipment requirements
A Google account and laptop are required to participate in the hands-on exercises. Related files will be sent via email at least three business days prior to May 3. 

CME Credits

CME credits are not available for presentations in this session. 

*Presenters and presentations are subject to change. 

All-Access Saturday pass

NEW: ARVO's All-Access Saturday pass grants you unlimited all-day access to your choice of Saturday sessions for just one registration fee. You have the freedom to move around to various presentations you are interested in within sessions and build your own individual schedule for the day. To sign up, simply add the All-Access Saturday pass to your registration for the 2025 Annual Meeting.

Registration for ARVO's All-Access Saturday is open.

* Breakfast and lunch included with registration