Search for: All records

Substandard and falsified (SF) pharmaceuticals account for an estimated 10% of the pharmaceutical supply chain in low- and middle-income countries (LMICs), where a lack of regulatory and laboratory resources limits the ability to conduct effective post-market surveillance and allows SF products to penetrate the supply chain. The Distributed Pharmaceutical Analysis Laboratory (DPAL) was established in 2014 to expand testing of pharmaceutical dosage forms sourced from LMICs; DPAL is an alliance of academic institutions throughout the United States and abroad that provides high quality, validated chemical analysis of pharmaceutical dosage forms sourced from partners in LMICs. Results from analysis are reported to relevant regulatory agencies and are used to inform purchasing decisions made by in-country stakeholders. As the DPAL program has expanded to testing more than 1000 pharmaceutical dosage forms annually, challenges have surfaced regarding data management and sample tracking. Here, we describe a pilot project between DPAL and ARTiFACTs that applies blockchain to organize and manage key data generated during the DPAL workflow, including a sample’s progress through the workflow, its physical location, provenance of metadata, and lab reputability. Recording time and date stamps with this data will create a permanent and verifiable chain-of-custody for samples. This secure, distributed ledger will be linked to an easy-to-use dashboard, allowing stakeholders to view results and experimental details for each sample in real time and verify the integrity of DPAL analysis data. Introducing this blockchain-based system as a pilot will allow us to test the technology with real users analyzing real samples. Feedback from users will be recorded and necessary adjustments will be made to the system before the implementation of blockchain across all DPAL sites. Anticipated benefits of implementing blockchain for managing DPAL data include efficient management for routing work, increasing throughput, creating a chain of custody for samples and their data in alignment with the distributed nature of DPAL, and using the analysis results to detect patterns of quality within and across brands of products and develop enhanced sampling techniques and best practices. 

Many cyber attack actions can be observed but the observables often exhibit intricate feature dependencies, non-homogeneity, and potential for rare yet critical samples. This work tests the ability to model and synthesize cyber intrusion alerts through Generative Adversarial Networks (GANs), which explore the feature space through reconciling between randomly generated samples and the given data that reflects a mixture of diverse attack behaviors. Through a comprehensive analysis using Jensen-Shannon Divergence (JSD), conditional and joint entropy, and mode drops and additions, we show that the Wasserstein-GAN with Gradient Penalty and Mutual Information (WGAN-GPMI) is more effective in learning to generate realistic alerts than models without Mutual Information constraints. The added Mutual Information constraint pushes the model to explore the feature space more thoroughly and increases the generation of low probability yet critical alert features. By mapping alerts to a set of attack stages it is shown that the output of these low probability alerts has a direct contextual meaning for cyber security analysts. Overall, our results show the promising novel use of GANs to learn from limited yet diverse intrusion alerts to generate synthetic ones that emulate critical dependencies, opening the door to data driven network threat models.