Email application materials to email@example.com
Healthcare Data Science: Discovering and Distributing the Latent Knowledge Embedded in Clinical Data.
Healthcare politics and economics is about hard tradeoffs between the cost, quality, and access to healthcare. The only way to improve all three is science and technology to advance the frontiers of practice. The world needs people like you who are ready to tackle complex problems in healthcare through data science and decision support solutions.
In Stanford's Center for Biomedical Informatics Research, you will have the opportunity to work in close collaboration with clinicians, scientists, and healthcare systems with access to deep clinical data warehouses (e.g., electronic medical records), broad population health data sources (e.g., national claims), and professional development resources like (grant) writing workshops and clinical shadowing experiences. Research topics range from machine learning, designing, and evaluating clinical decision support content to disintermediate scarce medical consultation resources, systematically identifying ineffective clinical processes, data-mining audit logs of electronic clinical activity, as well as more conventional outcomes research on the implications of physician practice against challenging issues in healthcare.
Specific near-term (funded) projects include:
(1) Developing prediction models from electronic medical records for cognitive impairment and dementia in elder patients. Developing explainable recommender models to predict diagnostic procedures for the evaluation of such cases.
(2) Identifying cohorts of patients in electronic medical records and national claims databases who received treatment for substance use who either did or did not continue with longer term treatment. Developing models to predict and identify candidate risk factors for such treatment retention using both structured clinical data and keyword concepts extracted through a natural language processing infrastructure.
The position will allow for exploration of additional research threads further tailored to the applicant's interests and career goals.
- The strongest applicants will have experience in one or more key interdisciplinary areas (not all are expected, that's the point of learning together):
- Computer Science or Informatics:
- Proficiency in programming and software development with a habit for robust unit testing. Our group mainly develops software in a Python + SQL environment with R for additional statistical analysis. For decision support prototype development, web-based user interface design and human-computer interaction testing experience will be valuable.
- Statistics and Mathematics:
- Machine learning (supervised and unsupervised) methodology and evaluation including discrimination vs. calibration measures and (hyper)parameter optimization through cross-validation. Observational research methods including interpreting multivariate regression, missing data imputation, propensity score matching, and bootstrap simulations.
- Biomedical / Healthcare Science:
- Understanding of clinical decision making processes, healthcare quality metrics, financial incentives, and decision support interfaces and pitfalls.
- Computer Science or Informatics:
- A PhD in a quantitative field with a strong programming and statistics background
- Track record of completed research projects
- Well-written, peer reviewed papers is expected
- Specific responsibilities and research projects will be tuned to the career goals, technical strengths, and interests of the applicant.
- Example research paper
- 2-3 references
- A brief career goal statement (that reflects alignment with the projects we would likely pursue together)