ABSTRACT
Medical artificial intelligence (AI) has tremendous potential to advance healthcare by supporting and contributing to the evidence-based practice of medicine, personalizing patient treatment, reducing costs, and improving both healthcare provider and patient experience. Unlocking this potential requires systematic, quantitative evaluation of the performance of medical AI models on large-scale, heterogeneous data capturing diverse patient populations. Here, to meet this need, we introduce MedPerf, an open platform for benchmarking AI models in the medical domain. MedPerf focuses on enabling federated evaluation of AI models, by securely distributing them to different facilities, such as healthcare organizations. This process of bringing the model to the data empowers each facility to assess and verify the performance of AI models in an efficient and human-supervised process, while prioritizing privacy. We describe the current challenges healthcare and AI communities face, the need for an open platform, the design philosophy of MedPerf, its current implementation status and real-world deployment, our roadmap and, importantly, the use of MedPerf with multiple international institutions within cloud-based technology and on-premises scenarios. Finally, we welcome new contributions by researchers and organizations to further strengthen MedPerf as an open benchmarking platform.
ABSTRACT
Online social network users may leave creative, subtle cues on their public profiles to communicate their motivations and interests to other network participants. This paper explores whether psychological predictions can be made about the motivations of social network users by identifying and analyzing these cues. Focusing on the domain of relationship seeking, we predicted that people using social networks for dating would reveal that they have a single relationship status as a method of eliciting contact from potential romantic others. Based on results from a pilot study (n = 20) supporting this hypothesis, we predicted that people attempting to attract users of the same religious background would report a religious affiliation along with a single relationship status. Using observational data from 150 Facebook profiles, results from a multivariate logistic regression suggest that people providing a religious affiliation were more likely to list themselves as single (a proxy for their interest in using the network to find romantic partners) than people who do not provide religious information. We discuss the implications for extracting psychological information from Facebook profiles. To our knowledge, this is the first study to suggest that information from publicly available online social networking profiles can be used to predict people's motivations for using social networks.
