Using Big Data to Compare the Effect of Localized Dengue Vector Control Interventions
Dengue virus is transmitted by the Aedes mosquitoes and causes over 96 million infections worldwide every year. The virus is expanding in geographic range and magnitude, and has created a significant global health concern. Several interventions exist to control dengue that target the vector population, but to-date, a lack of comprehensive empirical data has led to limited and conflicting evidence regarding their efficacy. Here, we supplemented data from the world’s largest dengue containment monitoring system with large-scale climate and mobility data to better understand the impact of several common containment activities, on the reproductive number of dengue, at a sub-city granularity.
Completed Research Projects
Fine-grained Dengue Forecasting using Telephone Triage Services
Thousands of lives are lost every year in developing countries for failing to detect epidemics early because of the lack of real-time disease surveillance data. We present results from a large-scale deployment of a telephone triage service as a basis for dengue forecasting in Pakistan. Our system uses statistical analysis of dengue-related phone calls to accurately forecast suspected dengue cases 2 to 3 weeks ahead of time at a subcity level (correlation of up to 0.93). Our system has been operational at scale in Pakistan for the past 3 years and has received more than 300,000 phone calls. The predictions from our system are widely disseminated to public health officials and form a critical part of active government strategies for dengue containment. Our work is the first to demonstrate, with significant empirical evidence, that an accurate, location-specific disease forecasting system can be built using analysis of call volume data from a public health hotline.
Read our paper (link)
We performed the first detailed algorithmic analysis of how Google Flu Trends can be used as a basis for building a fully automated system for early warning of epidemics. We explored the relative merits of three types of algorithms: normal distribution algorithms, Poisson distribution algorithms, and negative binomial distribution algorithms, and related our findings to changes in Internet penetration and population size of the regions for which Google Flu Trends provided data. Based on our work, we developed FluBreaks, an early warning system for flu epidemics using Google Flu Trends. We published our findings in Journal of Medical Internet Research (JMIR). Read our paper (link) Link to our system (link)
Dengue Early Epidemic Detection System for the Punjab Government
By leveraging a combination of supervised and unsupervised machine learning techniques, and using GPS tagged dengue patients and larvae sightings reports, and weather indicators as data sources, I developed a spatio-temporal early epidemic detection system to identify potential hotspot for dengue spread. The system to is used by the government officials on a daily basis to allocate work force to clean the high alert areas of potential dengue spread. Link to our system (link)
Characterizing Dengue Spread and Severity using Internet Media Sources
This project aims to build a system that automatically aims to characterize the spread and severity of the dengue disease at a fine-grained location granularity based on analyzing news reports from Internet media sources. Our system leverages a range of standard data mining and machine learning techniques to arrive at an accurate dengue severity measure for any given location. Based on a detailed analysis of news reports gathered from several leading dailies in Pakistan, we demonstrate the effectiveness of our system to accurately characterize the dengue spread and severity across different locations within Pakistan. Our finding were accepted as a poster in Proceedings of the 3rd ACM Symposium on Computing for Development(DEV). Read our paper (link) Currently, we are working on extending this system to cover Malaria disease in Pakistan.