In this Comment, mars huang, Akshay Chaudhari, Curt Langlotz, Nigam Shah, @syeung10, Matthew Lungren MD MPH discuss best practices when developing medical imaging models for emerging infectious diseases Stanford AIMI nature.com/articles/s4146…

Great review led by mars huang on lessons learned for medical imaging ML during COVID!

2. DrML: Diagnosing and Rectifying Vision Models using Language Session: Sat 12/3 1:00PM-2:30PM @ Room 388-390 Joint work with Haochen Zhang, mars huang, Jackson (Kuan-Chieh) Wang, James Zou, @syeung10

3. Adapting Pre-trained Vision Transformers from 2D to 3D through Weight Inflation Improves Medical Image Segmentation Session: Mon 11/28 4:40PM-5:25PM @ ML4H (Intercontinental New Orleans) Joint work with mars huang, Zhengping Zhou, Matthew Lungren MD MPH, @syeung10

Interested to find & fix error modes of image classifiers using text prompts? Our work to appear at ICLR23 shows when/how we can do this using text embeddings as proxy for image embeddings. Led by Yuhui Zhang w/ Haochen Zhang mars huang Jackson (Kuan-Chieh) Wang James Zou Code released!

This #systematicreview of deep learning models that leverage self-supervised learning for medical image classification tasks aggregates the collective knowledge of prior work, consolidates terminology, and offers implementation guidelines. nature.com/articles/s4174…

New systematic review and guidelines on self-supervised learning for medical imaging just dropped! nature.com/articles/s4174… Congrats to mars huang Akshay Chaudhari @syeung10 Anuj Pareek

Our new npj Digital Medicine systematic review covers SSL for medical imaging classification. We explore the methods, their benefits, and some future directions here! Great work co-led by mars huang + Anuj Pareek, and great to collaborate with Malte, Matthew Lungren MD MPH and @syeung10

Could Self-Supervised Learning Be a Game-Changer for Medical Image Classification? “self-supervised training may be a step on the path to a true foundation model for medical image classification” mars huang Akshay Chaudhari @syeung10 Anuj Pareek nature.com/articles/s4174…

Have videos of your tennis practice and wish you can put your own motion in 3D? 🎾 👟 🏋🏻 #CVPR2023 We present, NeMo, a 3D motion recovery method that is more accurate by leveraging information shared across multiple instances/repetitions! 👇🏻Resources in 🧵

Thanks for tweeting, @AK!  We’re super excited about the future of text-only vision model selection! 🙏 mars huang Jackson (Kuan-Chieh) Wang @cvpr @syeung10

Pls stop at #CVPR2023 poster *Tue AM 110* to learn about GC-KPL: a novel method for learning 3D human keypoints from point clouds w/o human labels. Project: cvpr2023.thecvf.com/virtual/2023/p… Joint work w/ awesome folks Alexander Gorban Jingwei Ji, Mahyar Najibi, Yin Zhou, Dragomir Anguelov, Waymo

A new systematic review evaluates the impact of self-supervised learning in medical image classification. Findings show improved model performance, especially in radiology. Combining different SSL strategies appears promising. nature.com/articles/s4174…

There's a lot of excitement around large language models (LLMs) for healthcare. But what's hype and what's real? In this paper, we review 84 such models to help health systems better understand and critically evaluate these technologies. Paper: nature.com/articles/s4174… (1/7)

GPT-4 for matching more clinical trials with patients🧐 Featured by AK on HuggingFace Daily Papers huggingface.co/papers/2308.02…

It was fun to summarize lessons learned from research in partnership with Stanford HAI, Center for Research on Foundation Models, Stanford Medicine for our clinical colleagues. We have to verify the claimed value propositions (hai.stanford.edu/news/how-found…) because they don't always pan out (hai.stanford.edu/news/how-well-…)!

Lots of hype around #LLMs in healthcare. What do clinicians really want from an #LLM? We asked them! Introducing #MedAlign, the first dataset of clinician-generated instructions + responses for EHRs 🏥🤖 📄Paper: arxiv.org/abs/2308.14089 🌐Website: medalign.stanford.edu

We're excited to introduce #INSPECT a large-scale ✨3D multimodal✨medical imaging dataset #NeurIPS2023 19,402 Stanford Medicine Patients 🩻23,248 CT Scans + 📄Paired Radiology Notes 📈Longitudinal EHRs 🩺Clinician-validated task labels #DataCentricAI #Multimodal #3Dimaging 1/

CC @stanfordAIMI @stanfordHAI @VinBrainAI Pierre Chambon mars huang Zhihong Chen Maya Varma steven truong Thomas Sounack JB Chu The Chuong

All the CheXpert Plus links in one place: --Dataset: aimi.stanford.edu/shared-datasets --Arxiv paper: arxiv.org/abs/2405.19538 --De-ID algorithm paper: academic.oup.com/jamia/article-… --Hugging Face De-ID algorithm: huggingface.co/StanfordAIMI/s… Stanford AIMI