MD/PhD candidate @VanderbiltMSTP doing translational data science @MASI_Lab. #DukeU '18

Joined October 2016
2 Photos and videos
9 Oct 2023
Poster is up at #MICCAI2023! We discuss learning strategies for integrating longitudinal Imaging EHRs. I will be at the 9:30 AM poster session (T-3-73) and then presenting at 4:30 PM (oral 7). Looking forward to my first MICCAI with @MASI_Lab!
2
13
1,645
24 Mar 2023
Someone throw the uworld 2ck qbank at this. Mine expired…
🚨 NEW Version of Glass AI Available 🚨 Glass AI 2.0 combines an LLM with a clinical knowledge database created by our clinician team. With clinician training, we create better DDx and Clinical Plan outputs than base LLMs. Try 2.0 now at glass.health/ai #medtwitter
2
587
Thomas Li retweeted
Replying to @LeonCai
@LeonCai handed off the GitHub Guru trophy for lab mentorship. Thank you @_thomasli for filling his very big shoes. I appreciate our great @VanderbiltMSTP / @VandyBME team players.
1
3
17
1,720
Thomas Li retweeted
We are very interested in interpreting images within the context of a patient's history, known biomarkers, and other available data. However, all too often, not every patient has every data type available. Recently, we looked at this problem in lung screening.
1
2
3
7 Sep 2022
Our work on vision transformers for longitudinal medical imaging is out as preprint! We leverage time intervals b/w repeated scans to improve on a cross-sectional approach. arxiv.org/abs/2209.01676 with @MASI_Lab @KXu63 @tangyc1183 @MaldonadoFabien @SandlerRhodes

1
3
10
2 May 2022
Great thread on understanding transformers from (roughly) first principles
Transformers are arguably the most impactful deep learning architecture from the last 5 yrs. In the next few threads, we’ll cover multi-head attention, GPT and BERT, Vision Transformer, and write these out in code. This thread β†’ understanding multi-head attention. 1/n
19 Mar 2022
So excited that my fiancΓ© Aiwei will be joining @VUMCMedicineRes!
6
1
53
16 May 2021
The breadth that @EyeOn_AI covers is amazing - research, startups, policy, ethics, verticals... list goes on
2
Thomas Li retweeted
Brand new PredictionHealth video hot and fresh out of the oven! πŸ•ΊπŸ©Ί πŸ€– So psyched to be saving clinicians hours a day and freeing them up to focus on their patients. This is just the beginning, too!
You didn't go into medicine to spend hours in the EHR. Let AI-assisted scribes do your documentation for you. Introducing PredictionHealth.
2
2
Thomas Li retweeted
We would like to introduce Lucky (right) with her best friend (Leia) to the MSTPet community! Photo credit goes to Lucky’s human @_thomasli (M2).
1
4
Thomas Li retweeted
Musings of an anonymous, pissed off virologist.
319
4,558
12,162
Thomas Li retweeted
23 Jan 2020
What company is the #1 market #EHR, makes >$3 billion/yr selling (aka gouging) health systems poor performing software, and now opposes @HHSGov proposal for hospitals to share medical data for their patients? This one: cnbc.com/2020/01/22/epic-ceo… by @chrissyfarr @CNBC
28
171
341
Thomas Li retweeted
6 Jan 2020
A prospective clinical trial of #AI to determine intraoperative #cancer in near real time using label-free imaging: impressive rapid, accurate results by @umichneuro @ToddCHollon @DanOrringerMD & collaborators @NatureMedicine nature.com/articles/s41591-0…
10
103
188
25 Sep 2019
First Apple now Amazon. Interesting that they built out the mobile app and rx delivery but stopped short on hiring doctors. Instead they contracted with an existing provider
SCOOP: Amazon just launched Amazon Care, virtual medical clinics for its own employees. That includes pharmacy, virtual visits, and in-home care. cnbc.com/2019/09/24/amazon-l…
Thomas Li retweeted
πŸ€–πŸ“’β€ΌοΈML Healthcare Researchers β€ΌοΈπŸ“’πŸ€– BIG medical data release thanks to @alistairewj @MIT @BIDMChealth MIMIC 2.0 >375k chest Xrays 𝙁𝙍𝙀𝙀 𝙏𝙀𝙓𝙏 π™π˜Όπ˜Ώπ™„π™Šπ™‡π™Šπ™‚π™” π™π™€π™‹π™Šπ™π™π™Ž Link: physionet.org/content/mimic-… cc: @jeremyphoward @jeremy_irvin16
19
359
998
Thomas Li retweeted
Soooo excited to see this out @LancetDigitalH - deep learning for medical image classification by doctors @Moorfields with NO coding experience using @googlecloud #DeepLearning #AutoML #OpenSource authors.elsevier.com/sd/arti…
10
136
352
11 Sep 2019
Sad that I JUST discovered this paper. What's the equivalent technique for understanding representations from audio, text, ECG, genome... data?!!
6 Mar 2019
Exploring Neural Networks with Activation Atlases -- A new Distill article by @shancarter, Zan Armstrong, @ludwigschubert, @enjalot, and @ch402. distill.pub/2019/activation-…
2