The healthcare sector has been confronted with rapidly rising healthcare costs and a shortage of medical staff. At the same time, the field of Artificial Intelligence (AI) has emerged as a promising area of research, offering potential benefits for healthcare. Despite the potential of AI to support healthcare, its widespread implementation, especially in healthcare, remains limited. One possible factor contributing to that is the lack of trust in AI algorithms among healthcare professionals. Previous studies have indicated that explainability plays a crucial role in establishing trust in AI systems. This study aims to explore trust in AI and its connection to explainability in a medical setting. A rapid review was conducted to provide an overview of the existing knowledge and research on trust and explainability. Building upon these insights, a dashboard interface was developed to present the output of an AI-based decision-support tool along with explanatory information, with the aim of enhancing explainability of the AI for healthcare professionals. To investigate the impact of the dashboard and its explanations on healthcare professionals, an exploratory case study was conducted. The study encompassed an assessment of participants’ trust in the AI system, their perception of its explainability, as well as their evaluations of perceived ease of use and perceived usefulness. The initial findings from the case study indicate a positive correlation between perceived explainability and trust in the AI system. Our preliminary findings suggest that enhancing the explainability of AI systems could increase trust among healthcare professionals. This may contribute to an increased acceptance and adoption of AI in healthcare. However, a more elaborate experiment with the dashboard is essential.
LINK
Purpose: The aims of this study were to investigate how a variety of research methods is commonly employed to study technology and practitioner cognition. User-interface issues with infusion pumps were selected as a case because of its relevance to patient safety. Methods: Starting from a Cognitive Systems Engineering perspective, we developed an Impact Flow Diagram showing the relationship of computer technology, cognition, practitioner behavior, and system failure in the area of medical infusion devices. We subsequently conducted a systematic literature review on user-interface issues with infusion pumps, categorized the studies in terms of methods employed, and noted the usability problems found with particular methods. Next, we assigned usability problems and related methods to the levels in the Impact Flow Diagram. Results: Most study methods used to find user interface issues with infusion pumps focused on observable behavior rather than on how artifacts shape cognition and collaboration. A concerted and theorydriven application of these methods when testing infusion pumps is lacking in the literature. Detailed analysis of one case study provided an illustration of how to apply the Impact Flow Diagram, as well as how the scope of analysis may be broadened to include organizational and regulatory factors. Conclusion: Research methods to uncover use problems with technology may be used in many ways, with many different foci. We advocate the adoption of an Impact Flow Diagram perspective rather than merely focusing on usability issues in isolation. Truly advancing patient safety requires the systematic adoption of a systems perspective viewing people and technology as an ensemble, also in the design of medical device technology.
Many healthcare professionals experience difficulties in discussing sexual health with their patients. The aim of this review was to synthesize results of studies on communication practices in interactions about sexual health in medical settings, to offer healthcare professionals suggestions on how to communicate about this topic. Veel zorgprofessionals ervaren problemen bij het bespreken van seksuele gezondheid met hun patiënten. Het doel van deze review was een synthese te presenteren van studies naar communicatiepraktijken in interactie over seksuele gezondheid in medische settings, om zorgprofessionals handreikingen te bieden voor communicatie over dit thema.
The modern economy is largely data-driven and relies on the processing and sharing of data across organizations as a key contributor to its success. At the same time, the value, amount, and sensitivity of processed data is steadily increasing, making it a major target of cyber-attacks. A large fraction of the many reported data breaches happened in the healthcare sector, mostly affecting privacy-sensitive data such as medical records and other patient data. This puts data security technologies as a priority item on the agenda of many healthcare organizations, such as of the Dutch health insurance company Centraal Ziekenfonds (CZ). In particular when it comes to sharing data securely, practical data protection technologies are lacking as they mostly focus on securing the link between two organizations while being completely oblivious of what is happening with the data after sharing. For CZ, searchable encryption (SE) technologies that allow to share data in encrypted form, while enabling the private search on this encrypted data without the need to decrypt, are of particular interest. Unfortunately, existing efficient SE schemes completely leak the access pattern (= pattern of encrypted search results, e.g. identifiers of retrieved items) and the search pattern (= pattern of search queries, e.g. frequency of same queries), making them susceptible to leakage-abuse attacks that exploit this leakage to recover what has been queried for and/or (parts of) the shared data itself. The SHARE project will investigate ways to reduce the leakage in searchable encryption in order to mitigate the impact of leakage-abuse attacks while keeping the performance-level high enough for practical use. Concretely, we propose the construction of SE schemes that allow the leakage to be modeled as a statistic released on the queries and shared dataset in terms of ε-differential privacy, a well-established notion that informally says that, after observing the statistic, you learn approximately (determined by the ε-parameter) the same amount of information about an individual data item or query as if the item was not present in the dataset or the query has not been performed. Naturally, such an approach will produce false positives and negatives in the querying process, affecting the scheme’s performance. By calibrating the ε-parameter, we can achieve various leakage-performance trade-offs tailored to the needs of specific applications. SHARE will explore the idea of differentially-private leakage on different parts of SE with different search capabilities, starting with exact-keyword-match SE schemes with differentially-private leakage on the access pattern only, up to schemes with differentially-private leakage on the access and search pattern as well as on the shared dataset itself, allowing for more expressive query types like fuzzy match, range, or substring queries. SHARE comes with an attack lab in which we investigate existing and new types of leakage-abuse attacks to assess the mitigation-potential of our proposed combination of differential privacy with cryptographic guarantees in searchable encryption. To stimulate commercial exploitation of SHARE-results, our consortium partners CZ and TNO will take the lead on applying and evaluating our envisioned technologies in various healthcare use-cases.
The goal of UPIN is to develop and evaluate a scalable distributed system that enables users to cryptographically verify and easily control the paths through which their data travels through an inter-domain network like the Internet, both in terms of router-to-router hops as well as in terms of router attributes (e.g., their location, operator, security level, and manufacturer). UPIN will thus provide the solution to a very relevant and current problem, namely that it is becoming increasingly opaque for users on the Internet who processes their data (e.g., in terms of service providers their data passes through as well as what jurisdictions apply) and that they have no control over how it is being routed. This is a risk for people’s privacy (e.g., a malicious network compromising a user’s data) as well as for their safety (e.g., an untrusted network disrupting a remote surgery). Motivating examples in which (sensitive) user data typically travels across the Internet without user awareness or control are: - Internet of Things for consumers: sensors such as sleep trackers and light switches that collect information about a user’s physical environment and send it across the Internet to remote services for analysis. - Medical records: health care providers requiring medical information (e.g., health records of patients or remote surgery telemetry) to travel between medical institutions according to specified agreements. - Intelligent transport systems: communication plays a crucial role in future autonomous transportation systems, for instance to avoid freight drones colliding or to ensure smooth passing of trucks through busy urban areas. The UPIN project is novel in three ways: 1. UPIN gives users the ability to control and verify the path that their data takes through the network all the way to the destination endpoint, both in terms of hops and attributes of routers traversed. UPIN accomplishes this by adding and improving remote attestation techniques for on-path routers to existing path verification mechanisms, and by adopting and further developing in-packet path selection directives for control. 2. We develop and simulate data and control plane protocols and router extensions to include the UPIN system in inter-domain networking systems such as IP (e.g., using BGP and segment routing) and emerging systems such as SCION and RINA. 3. We evaluate the scalability and performance of the UPIN system using a multi-site testbed of open programmable P4 routers, which is necessary because UPIN requires novel packet processing functions in the data plane. We validate the system using the earlier motivating examples as use cases. The impact we target is: - Increased trust from users (individuals and organizations) in network services because they are able to verify how their data travels through the network to the destination endpoint and because the UPIN APIs enable novel applications that use these network functions. - More empowered users because they are able to control how their data travels through inter-domain networks, which increases self-determination, both at the level of individual users as well as at the societal level.