From 5000+ participants in the @NIH@AllofUsResearch:
U-shaped relationship between HDL and AMD risk, such that both high and low HDL are significantly associated with increased AMD risk. @jimmyschen@SallyLiuBaxter@EskoJeff@aao_ophth paper: https://t.co/hYJR7NMRVw
Chen et al published an open-source annotated glaucoma medication dataset derived from clinical notes in the EHR in @ARVOtvst, an #openaccess journal for the #visionscience community. So, annotated free-text clinical notes can be de-identified for upload as an open data set.
So true, a humble reminder for all of us in #informatics and #DataScience to continue working towards equitable healthcare access for EVERYONE #healthequity
Honored to present my thoughts at our @UCSF Clinical Informatics Fellowship graduation today! Parting thoughts: "missing data" is too easy of a phrase. In fact, I think the MOST important data in the EHR isn't actually there…
Using limited data, we've trained generative adversarial networks (GANs) to synthesize medical images, which can be used to train convolutional neural networks (CNNs) that achieve expert-level performance for detection of rare diseases.
1/16
https://t.co/x0RDSnjJgt
Should we be training neural networks differently?
Cool things can happen when you work wtih experts from other fields. Recent paper by @caseyeye resident Adam Hanif and recent @Northeastern PhD graduate Ilkay Yildiz.
@NEIDirector@rvpchan@qtimlab
https://t.co/2YHyOhci5o
Large, diverse datasets are necessary to do good #DataScience research! Data sharing could seriously accelerate innovations in vision research. #Ophthalmology#ophthotwitter
This will hopefully inspire more medical researchers to share their data! Link to the GitHub repo: https://t.co/m24q0peB31 #ophthalmology#MedTwitter#DataScience
Can incorporation of an #AI-based vascular severity score into a #RetinopathyofPrematurity (#ROP) risk model improve specificity and STILL have 100% sensitivity? You bet, and using just a single retinal fundus image acquired at 32 weeks PMA!
1/10
https://t.co/ng4VZqdlhP
I hope trainees in #Ophthalmology, #Optometry, and vision research will truly appreciate the impact of #vision on quality of life - please consider Retweeting.
From @JAMAOphth editorial on @NatEyeInstitute Strategic Plan, here's an anecdote that gave me insight on this...
1/7
Clinicians spend a lot of time writing and reviewing documentation in the EHR, which results in significant #burnout. We still have a lot of work to do as a community working towards EHR designs that better support clinician workflows!
From my own work:
#EHR requires significant documentation by clinicians. So what % of these notes are actually reviewed during a patient visit?
Answer from audit log data analysis: only 2.6%.
We need better approaches to EHR design & documentation. @JAMIA_Open@AMIAinformatics
From my own work: International Classification of Retinopathy of Prematurity, 3rd Edition paper is published @AAOjournal. Will be a standard reference for clinicians & researchers. #ophthotwitter#medtwitter#neonatology
Available free for 50 days: https://t.co/2HEvaDARjZ
1/5
Thank you for folks who have shared or commented on our paper. I know the paper is being used by some to dunk on Epic. Rather than piling on, I want to provide a clear-eyed view of what we found, what it means, and what I would suggest to Epic (& other model devs) going forward.
awesome work by @jimmyschen future resident at @ShileyEye, @OHSUSOM med student, and @RPB_org med student research fellowship grant recipient at @CaseyEye. reiterating fundamental principles of AI in ophthalmology with a potential for global impact.
https://t.co/KtJIgnWIpe