Myura Nagendran

Myura Nagendran

1 previous affiliation
3Machine learning papers tracked by Lacuna
Is this you?

Research Overview

Myura Nagendran focuses on the intersection of explainable artificial intelligence, clinical decision support, and human-computer interaction in high-stakes medical environments. Early work investigated how intensive care physicians interact with AI suggestions using high-fidelity simulations and eye-tracking technology. Current research explores the cognitive load and behavioral dynamics of doctors utilizing explainable reinforcement learning systems to ensure patient safety.

Research Trajectory

In 2023, Nagendran led research investigating how medical professionals interact with artificial intelligence in critical care scenarios. A primary focus of this period was the broader area of Bridging the Clinical Translation Gap for Machine Learning, addressing the challenge that AI models often perform well in laboratories but fail to integrate effectively into real-world clinical workflows. As the first author of Eye-tracking of clinician behaviour with explainable AI decision support: a high-fidelity simulation study, Nagendran examined intensive care doctors performing prescription dosing tasks. By utilizing eye-tracking glasses in a high-fidelity simulation suite, this work evaluated how clinicians respond to both safe and intentionally unsafe (hallucinatory) AI suggestions, contributing to the research direction of Ecological Validity in High Stakes Human Behavior Research.

Building on these simulation methods, Nagendran's subsequent first-author work focused specifically on reinforcement learning applications in healthcare. The paper Interaction of doctors with explainable RL decision support via behavioural readouts of eye-tracking explored how explainable reinforcement learning (XRL) might mitigate safety concerns that have historically slowed the deployment of dynamic AI tools in medicine. This research intersects with the theme of Learning Optimal Sequences of Medical Interventions, analyzing the cognitive attention doctors devote to XRL outputs during critical medication dosing scenarios.

In 2024, Nagendran continued to advance this methodology as a second author on Evaluating the impact of explainable RL on physician decision-making in high-fidelity simulations: insights from eye-tracking metrics. This work further quantified the behavioral and cognitive impacts of XRL on physician decision-making. By leveraging eye-tracking metrics to understand internal mental effort, this collaborative research aligns with the broader area of Physiological Proxies for Real-Time Cognitive Load Estimation.

Current Directions

Recent work investigates the cognitive demands placed on clinicians by transparent AI systems. The most recent papers in the evidence address the research direction of Managing Cognitive Load in Explainable AI, exploring how detailed algorithmic explanations affect human decision-making under pressure. By continuously measuring behavioral readouts like eye-tracking during high-stakes simulations, Nagendran's ongoing research trajectory also informs the broader area of Longitudinal Dynamics of Human AI Trust, seeking to understand how medical experts calibrate their reliance on AI teammates over time and across varying levels of system reliability.

Profile updated

See something wrong? Report a mistake