Skip to content Skip to navigation

Complex Data and Deep Learning for Disease Outbreak Prediction, Stanford (1/1/2017 - 1/1/2018)

Disease outbreaks are not easily predicted because they occur only when multiple factors trigger the rapid spread of disease. Key factors can often be identified, e.g., excess rainfall leading to outbreaks of Rift Valley fever virus (RVFV)1,2, but the complex circumstances that lead to outbreaks remain elusive for several reasons. First, gathering varied datasets (climatic, genetics, demographic, historical, and behavioral) is time consuming and expensive. Second, the computing capabilities to mine and analyze such varied and complex datasets has not been available until recently. In this application using RVFV as a case study, we propose to model the interplay between vectors, livestock, wildlife, climate, and humans. Large historic and modern datasets are accessible to the key investigators and will be aggregated into a repository. In collaboration with an industry partner, we will then apply machine learning to construct models for inference and prediction of RVFV outbreaks. We achieve broad applicability by separating data gathering from deep learning and execution. As such, once data curation and conversion of a dataset has been completed, one can take advantage of deep learning in the absence of a computer expert. As deep learning makes few assumptions about the data, this approach is transferable to other outbreak scenarios and diseases.
Rift Valley fever (RVF) is a deadly vector-borne disease that infects livestock and humans3–6. Transmitted via mosquitoes to livestock7, it can decimate entire herds and cause catastrophic economic hardship8. Humans are exposed via vector and animal transmission: animal husbandry, slaughter and butchery, and ingesting diseased meat, milk, and blood9,10. RVFV is endemic throughout much of Africa, but has recently caused outbreaks the Middle East11,12 and has significant potential to spread to the EU and USA, where all the necessary vectors and hosts to allow transmission are present13–15. Our proposal initiates a new collaboration between faculty who are uniquely suited to investigate RVF and use DL and diverse datasets to predict RVF outbreaks. Dr. LaBeaud brings clinical expertise in RVF transmission and epidemiology3–6,16,17. Dr. Seetah brings expertise in historical climate, meat processing as ‘social and economic’ practice, and is a trained butcher. We collaborate with Dr. Kumm, CEO of insightAI, to adapt their GPU-based deep-learning platform to “learn” what triggers RVF outbreaks. All three have worked in Kenya and have established in-country networks. In addition, the team is in discussion with IBM who has recently acquired, providing access to a massive database of past climate. This transformative, cross-disciplinary research project meets all CIGH priority areas: climate change and global health, new solutions to improve health care delivery, new interdisciplinary collaborations among faculty, and is a high-impact, high-risk project that lends itself to implementation among stakeholders in both endemic and at risk regions of the world.




  • Krish Seetah, Assistant Professor of Anthropology, Stanford

Principal Investigator:

Angelle Desiree LaBeaud

Current Research Interests: 
Arthropod-borne viruses are emerging and re-emerging infections that are spreading throughout the world. Our laboratory investigates the epidemiology of arboviral infections, focusing on the burden of disease and the long-term complications on human health. In particular, Dr. LaBeaud investigates dengue, chikungunya, and Rift Valley fever viruses in Kenya, where outbreaks cause fever, arthritis, retinitis, encephalitis, and hemorrhagic fever. Our main research questions focus on the risk factors for arboviral infections, the development of diagnostic tests that can be administered in the...
Read more