Mobile health (mHealth) has emerged as a transformative solution to enhance healthcare accessibility and affordability, particularly in resource-constrained regions and low-to-middle-income countries. mHealth leverages mobile platforms to improve healthcare accessibility, addressing radiologist shortages in low-resource settings by enabling remote diagnosis and consultation through mobile devices. Mobile phones allow healthcare workers to transmit radiographic images, such as chest X-rays (CXR), to specialists or AI-driven models for interpretation. However, AI-based diagnosis using CXR images shared via apps like WhatsApp suffers from reduced predictability and explainability due to compression artifacts, and there is a lack of datasets to systematically study these challenges. To address this, we introduce CheXwhatsApp, a dataset of 141,804 paired original and WhatsApp-compressed CXR images. We present a benchmarking study which shows the dataset improves prediction stability and explainability of state-of-the-art models by up to 80%, while also enhancing localization performance. CheXwhatsApp is open-sourced to support advancements in mHealth applications for CXR analysis1.

A time-varying graph (TVG) can model useful practical scenarios such as intermittent contact between nodes. High degree nodes in such networks can act as central nodes for efficient implementation of different applications such as information dissemination. In this paper, we propose a distributed algorithm for finding a low cardinality set of high degree nodes in a time-varying graph. The algorithm efficiently exploits the overlap in coverage between nodes to reduce the size of the set and finds a small subset of all high degree nodes in the network while still maintaining almost the same coverage as the set of all high degree nodes. It also finds temporal paths from all nodes to the high degree nodes found which can be used for routing information to the nodes.