AMAngela Mastrianni

Angela Mastrianni

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Angela Mastrianni has a total of 35 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Designing an AI Speculative Facilitator for Healthcare Futures: From Prototype to Participatory Deployment

    by Katerina Andreadis, Angela Mastrianni, Lynn Xu, Veronica Alfaro Arias, Devin Mann, Katharine Lawrence

  • Open Access

    Evaluating the Evaluator: LLM Evaluators for Patient-Facing AI Agents

    by Angela Mastrianni, Katerina Andreadis, Ji Chen, Danissa V Rodriguez, Devin Mann


  • Open Access

    Lessons Learned from Deploying a GenAI-Powered Tool for Remote Hypertension Monitoring in a Large Academic Medical Center

    by Danissa V Rodriguez, Angela Mastrianni, Katerina Andreadis, Lynn Xu, Javier Gonzalez, Ji Chen, Devin Mann


  • Open Access

    Integrating Generative AI into Remote Patient Monitoring: Development and Deployment of the AI Brain, a GenAI-Enhanced Hypertension Management System at NYU Langone Health (Preprint)

    by Danissa Rodriguez, Katerina Andreadis, Angela Mastrianni, Ji Chen, Javier Gonzalez, Chelsea Twan, Lynn Xu, Devin Mann

    BACKGROUND NYU Langone Health (NYULH) operates one of the largest remote patient monitoring (RPM) programs in the United States. Its hypertension management initiative (NYULH RPM HTN) supports approximately 4,500 patients monthly and captured over 100,000 remote blood pressure (BP) readings in 2024. Despite its benefits, the program faces real-world challenges, including patient disengagement, device usability issues, and clinician burden from high data volume. Generative AI (GenAI), particularly large language models (LLMs), offers opportunities to enhance patient engagement and streamline clinical workflows through personalized conversational interfaces such as chatbots and its data summarization capabilities. OBJECTIVE To explore the feasibility of using GenAI to support RPM, we developed the AI Brain, an electronic health record (EHR)–integrated GenAI layer to support RPM for hypertension management. AI Brain includes a patient-facing agent, chatbot designed to support engagement and blood pressure (BP) adherence, as well as a clinician-facing agent that generates smart content for EHR documentation and drafts patient messages. This study was conducted at NYULH, an academic medical center, providing a unique setting to evaluate the tool within a large-scale hypertension RPM program and to assess its impact on patient engagement, data interpretation, and clinical workflow efficiency. METHODS Our multidisciplinary team—comprising researchers, software engineers/architects, UX designers, and physicians—developed the AI Brain using a user-centered design approach and agile software development methods. We established patient and clinician advisory committees and conducted workshops during the formative phase to understand workflows and co-design solutions in collaboration with stakeholders. This was followed by a software development cycle that engaged advisory committee members at each stage to ensure the tool met user needs. Implementation considerations included usability, data privacy, clinical integration, and alignment with existing RPM processes. RESULTS The evaluation of the AI Brain demonstrated feasibility for integration into an established RPM infrastructure. Early observations suggest that the patient-facing agent showed potential to address common engagement barriers, including missed blood pressure submissions and device-related challenges. The clinician-facing agent supported care teams by summarizing key patient trends and reducing manual data review burden. Moreover, structured survey results indicated positive acceptability and perceived usefulness of GenAI-generated content. Security evaluations further demonstrated robust safeguards and reliable system performance. CONCLUSIONS GenAI represents a promising approach to enhancing RPM, as demonstrated by its evaluation and adaptation within the NYULH hypertension management program. We described our development process and showed that, based on our evaluation, thoughtfully designed and integrated GenAI tools may help bridge gaps in patient workflows in terms of engagement and adherence as well as support clinical workflow to reduce data analysis and data summarization. Further evaluation is needed to assess long-term clinical outcomes, patient trust, and scalability in real-world healthcare settings.

  • Open Access

    AI-Enhanced Sensemaking: Exploring the Design of a Generative AI-Based Assistant to Support Genetic Professionals

    by Angela Mastrianni, Hope Twede, Aleksandra Sarcevic, Jeremiah Wander, Christina Austin-Tse, Scott Saponas, Heidi Rehm, Ashley Mae Conard, Amanda K. Hall

    Generative AI has the potential to transform knowledge work, but further research is needed to understand how knowledge workers envision using and interacting with generative AI. We investigate the development of generative AI tools to support domain experts in knowledge work, examining task delegation and the design of human–AI interactions. Our research focused on designing a generative AI assistant to aid genetic professionals in analyzing whole genome sequences (WGS) and other clinical data for rare disease diagnosis. Through interviews with 17 genetics professionals, we identified current challenges in WGS analysis. We then conducted co-design sessions with six genetics professionals to determine tasks that could be supported by an AI assistant and considerations for designing interactions with the AI assistant. From our findings, we identified sensemaking as both a current challenge in WGS analysis and a process that could be supported by AI. We contribute an understanding of how domain experts envision interacting with generative AI in their knowledge work, a detailed empirical study of WGS analysis, and three design considerations for using generative AI to support domain experts in sensemaking during knowledge work.


  • Open Access

    To Recommend or Not to Recommend: Designing and Evaluating AI-Enabled Decision Support for Time-Critical Medical Events

    by Angela Mastrianni, Mary Suhyun Kim, Travis M Sullivan, Genevieve Jayne Sippel, Randall S Burd, Krzysztof Z. Gajos, Aleksandra Sarcevic

    AI-enabled decision-support systems aim to help medical providers rapidly make decisions with limited information during medical emergencies. A critical challenge in developing these systems is supporting providers in interpreting the system output to make optimal treatment decisions. In this study, we designed and evaluated an AI-enabled decision-support system to aid providers in treating patients with traumatic injuries. We first conducted user research with physicians to identify and design information types and AI outputs for a decision-support display. We then conducted an online experiment with 35 medical providers from six health systems to evaluate two human-AI interaction strategies: (1) AI information synthesis and (2) AI information and recommendations. We found that providers were more likely to make correct decisions when AI information and recommendations were provided compared to receiving no AI support. We also identified two socio-technical barriers to providing AI recommendations during time-critical medical events: (1) an accuracy-time trade-off in providing recommendations and (2) polarizing perceptions of recommendations between providers. We discuss three implications for developing AI-enabled decision support used in time-critical events, contributing to the limited research on human-AI interaction in this context.


  • Open Access

    Beyond Decision Making: Considering Collaboration and Agency in the Design of AI-Based Decision-Support Systems for Fast-Response Medical Teams

    by Angela Mastrianni, Paige Kmetz-Cutrone, Kathryn Chang, Jonathan Y Stein, Aleksandra Sarcevic

    In addition to aiding decision making, AI-based clinical decision-support systems may need to consider and support provider agency and collaboration between medical providers. We analyzed collaboration and agency within fast-response teams, identifying implications for designing decision-support systems that not only facilitate decision making, but also collaboration and agency. Using an Actor-Network Theory approach, we reviewed videos of 12 pediatric trauma resuscitations and conducted a secondary analysis of 27 interviews with trauma team members. We identified actants in trauma resuscitation, shifts in agency that can occur within fast-response teams during medical emergencies, and factors considered by providers when envisioning the design of decision-support systems. From our analysis, we propose implications for existing human-AI interaction guidelines when designing AI systems for fast-response medical teams. We also highlight parallels between the introduction of clinical practice guidelines and the introduction of AI-based decision-support systems, suggesting that these systems may influence the training and ''clinical gaze'' of providers.


  • Designing computerized support for identifying and treating patients with life-threatening injuries during medical emergencies

    by Angela Xaviera Mastrianni, Aleksandra Sarcevic

    During medical emergencies, clinical teams evaluate and treat patients with life-threatening injuries. Clinical decision-support systems (CDSSs) can assist clinicians in evaluating and treating patients, helping to avoid misdiagnoses, delays, and errors. However, human-computer interaction issues, such as data entry barriers and poor workflow integration, often limit adoption of these systems. While advances in artificial intelligence (AI) can improve the accuracy of CDSSs, integrating AI within these systems can lead to additional HCI challenges. This dissertation focuses on developing computerized support for identifying and treating patients with severe blood loss or traumatic brain injury, two of the most frequent causes of preventable deaths during trauma resuscitation. The research has two aims: (1) design features to capture the data needed for identifying patients with severe blood loss or traumatic brain injury and (2) present decision support to the medical team. For the first aim, we developed features on a digital checklist to obtain the patient context information needed to identify patients at risk for severe blood loss or traumatic brain injury. We also designed and evaluated alerts to prompt timely documentation of critical information and examined automatic approaches to capturing data. For the second aim, we investigated different strategies for communicating information during time-critical medical events to support decision-making. We also examined collaboration, agency, and decision-making within teams responding to medical emergencies and the implications for designing CDSSs used by these teams.

  • Open Access

    Supporting Awareness of Dynamic Data: Approaches to Designing and Capturing Data within Interactive Clinical Checklists

    by Angela Mastrianni, Aleksandra Sarcevic, Hua Cui, Megan A. Krentsa, Travis M Sullivan, Issa Zakeri, Ivan Marsic, Randall S. Burd

    Automatically integrating data within interactive clinical checklists allows for enhanced dynamic displays, while also providing information needed for checklist adaptation to the context of the medical event. In this mixed-methods study, we used user-centered design sessions with clinicians to design a checklist interface that automatically captures and displays dynamic patient data. We compared the manual and automatic checklist versions during video-guided simulation sessions, evaluating the effects of automatic capture on clinicians’ interactions with dynamic data and their situation awareness. Despite clinicians’ concerns that automatic data capture would affect situation awareness, we found no significant difference in awareness scores. Participants preferred the automatic version, highlighting its improved accuracy and completeness. From our findings, we propose a framework for capturing dynamic data and designing dynamic data interfaces within interactive checklists. We conclude by discussing barriers and design opportunities for supporting awareness of data trends through checklists.