Latest information on COVID-19

Lab News

Jun June 2021

Check out the interview of Dr. Hernandez-Boussard by Russ Altman for Stanford Engineering’s The Future of Everything podcast, here.

Mar March 2021

We are now hiring postdoctoral fellows and staff to work on a new medical informatics project that will include NLP and machine/deep learning predictive analytics. Contact if if you are interested!

Feb February 09 Tue 2021

Interested in Deploying AI at point of Care? Tune into our overview of Bringing AI into Healthcare Safey and Effectively and register for the Coursera Course

Nov November 05 Thu 2020

Dr. Hernandez-Boussard discusses the application of machine learning and natural language processing (NLP) to Electronic Health Record (EHR) data at an NCI webinar. Check out the YouTube video here.

Jan January 29 Wed 2020

Stanford Strong. Diversity & Inclusion Week 2020 #BoussardLab

Recent Publications

Learning from past respiratory failure patients to triage COVID-19 patient ventilator needs: A multi-institutional study

Unlike well-established diseases that base clinical care on randomized trials, past experiences, and training, prognosis in COVID19 relies on a weaker foundation. Knowledge from other respiratory failure diseases may inform clinical decisions in this novel disease. This retrospective multicenter study trained machine learning (ML) models on patients hospitalized with COVID-like diseases (CLD) to predict invasive mechanical ventilation (IMV) within 48h in COVID-19 patients. ML models were trained on CLD patients at Stanford Hospital Alliance (SHA) and validated on hospitalized COVID-19 patients at both SHA and Intermountain Healthcare. In COVID-19 patients, the best models achieved an AUC of 0.77 at SHA and 0.65 at Intermountain. The performance of prediction models demonstrate high specificity and can be used as a triage tool at point of care. Read our paper.

Development and evaluation of novel ophthalmology domain-specific neural word embeddings to predict visual prognosis

In ophthalmology, a major challenge to the use of EHR to develop predictive algorithms is the inability to incorporate the wealth of information sequestered within the clinical free-text, and the use of domain-specific neural word embeddings may provide one solution. Here, we trained and evaluated novel word embeddings (WEs) specific to ophthalmology, using text corpora from published literature and electronic health records (EHR). We found that using ophthalmology domain-specific WEs improved performance in ophthalmology-related clinical prediction compared to general WEs. Deep learning models using clinical notes as inputs can predict the prognosis of visually impaired patients. This work provides a framework to improve predictive models using domain-specific WEs.