Rapidly scaling screening, testing and quarantine has shown to be an effective strategy to combat the COVID-19 pandemic. We consider the application of deep learning techniques to distinguish individuals with COVID from non-COVID by using data acquirable from a phone. Using cough and context (symptoms and meta-data) represent such a promising approach. Several independent works in this direction have shown promising results. However, none of them report performance across clinically relevant data splits. Specifically, the performance where the development and test sets are split in time (retrospective validation) and across sites (broad validation). Although there is meaningful generalization across these splits the performance significantly varies (up to 0.1 AUC score). In addition, we study the performance of symptomatic and asymptomatic individuals across these three splits. Finally, we show that our model focuses on meaningful features of the input, cough bouts for cough and relevant symptoms for context.
Testing capacity for COVID-19 remains a challenge globally due to the lack of adequate supplies, trained personnel, and sample-processing equipment. These problems are even more acute in rural and underdeveloped regions. We demonstrate that solicited-cough sounds collected over a phone, when analysed by our AI model, have statistically significant signal indicative of COVID-19 status (AUC 0.72, t-test,p <0.01,95% CI 0.61-0.83). This holds true for asymptomatic patients as well. Towards this, we collect the largest known(to date) dataset of microbiologically confirmed COVID-19 cough sounds from 3,621 individuals. When used in a triaging step within an overall testing protocol, by enabling risk-stratification of individuals before confirmatory tests, our tool can increase the testing capacity of a healthcare system by 43% at disease prevalence of 5%, without additional supplies, trained personnel, or physical infrastructure.
Nearly 100 million families across the world rely on cotton farming for their livelihood. Cotton is particularly vulnerable to pest attacks, leading to overuse of pesticides, lost income for farmers, and in some cases farmer suicides. We address this problem by presenting a new solution for pesticide management that uses deep learning, smartphone cameras, inexpensive pest traps, existing digital pipelines, and agricultural extension-worker programs. Although generic, the platform is specifically designed to assist smallholder farmers in the developing world. In addition to outlining the solution, we consider the set of unique constraints this context places on it: data diversity, annotation challenges, shortcomings with traditional evaluation metrics, computing on low-resource devices, and deployment through intermediaries. This paper summarizes key lessons learned while developing and deploying the proposed solution. Such lessons may be applicable to other teams interested in building AI solutions for global development.