AI for Health: Computational Infrastructure Needs
This form is designed to collect insights on how AI is being used in healthcare research and related computational infrastructure needs for such AI-enabled healthcare research applications.

This is in preparation for a grant proposal to the NSF MidScale Infrastructure program for a project focused on building a multi-site AI for healthcare testbed. This project is a joint effort among Univ of Delaware, Univ of Michigan, and USC.

Note to U-M researchers; the survey is *not* intended to address infrastructure within just Michigan Medicine.  

We aim to understand computational infrastructure needs and the proposed testbed's benefits. Your responses will help us build a community of individuals working in this space. 

If you know others who might be interested, please forward this link: URL (OR https://tinyurl.com/AI4HealthSurvey)

Thank you for your time!
Sign in to Google to save your progress. Learn more
Email *

First Name

 *
Last Name *

Affiliation (Organization Name)

 *
Affiliation Type *
Role *
Required
How do you use AI in your research, specifically in healthcare? (e.g., I develop AI models for diagnostic decision making, or I use LLMs to label data) *
What current AI/ML toolkits do you use (e.g., PyTorch, SciKit-learn, AutoML)? 
*
What data modalities do you use in your AI research (e.g., fMRI, physiological waveforms, tabular EHR, clinical notes, genomics)?
What are the biggest challenges or gaps with respect to computational infrastructure that you have encountered?
(e.g., privacy concerns, inability to share data, lack of GPUs, lack of data storage, lack of memory, lack of network bandwidth) *
Are there any data sharing issues or data privacy considerations in your use of AI/ML (e.g., how does HIPAA impact your work)? If so, describe them. *
Please clarify your need (or use) for multi-site support in your use of AI/ML.
Do you require data from other sites for training models?  If so, what challenges have you encountered in leveraging data across sites? *
Have you had any experiences with existing ML testbeds such as DeltaAI, NAIRR, or NDIF, or any other ML testbeds? 
If so, please comment on your positive and (if any) negative experiences in using them. *
Would you be willing to sign a letter attesting to the value and interest in using a secure multi-site platform for conducting AI/ML research in healthcare if it were made available to you?
(We will of course provide more details on the proposed ML testbed support). *
A copy of your responses will be emailed to the address you provided.
Submit
Clear form
Never submit passwords through Google Forms.
reCAPTCHA
PrivacyTerms
This form was created inside of University of Delaware.

Does this form look suspicious? Report

Google Forms