Annie Desmarais

Universite de Montreal, Canada

Title: Human-Centered AI in Neuropsychological and Mental Health Practice: Design, Ethics, and Clinician Acceptability

---

Biography

Annie Desmarais is a licensed neuropsychologist and postdoctoral researcher specializing in the integration of artificial intelligence into neuropsychology and mental health practice. She holds a Ph.D. in clinical neuropsychology from Université Laval (Canada), completed in collaboration with the Université de Liège (Belgium). She is currently a postdoctoral researcher at the Université de Montréal, affiliated with the IVADO Center of Excellence in Artificial Intelligence. Her research focuses on human-centered and ethical AI approaches to clinical assessment, decision support, and precision psychiatry. She is also the founder and CEO of Flow Factor, where she leads the development of Flow, a clinical AI platform co-designed with clinicians and healthcare institutions in Canada and Europe. Her work aims to advance responsible AI adoption while preserving clinical judgment and professional autonomy.

Abstract

Artificial intelligence (AI) is increasingly integrated into mental health practice, yet its clinical adoption remains limited by ethical concerns, questions of professional responsibility, and clinician acceptability. Beyond technical performance, the successful integration of AI in neurology and neuropsychology depends on alignment with clinical reasoning processes, ethical standards, and professional identities. The objective of this presentation is to examine these issues through a clinician-centered reflection on the design and implementation of Flow, an AI-enabled clinical platform developed in close collaboration with mental health professionals.

This work adopts a conceptual and practice-informed methodology grounded in participatory design and ethical-by-design principles. Rather than reporting empirical outcomes, the presentation draws on iterative co-design processes with neuropsychologists, neurologists, and mental health clinicians, focusing on real-world clinical workflows, assessment practices, and decision-making contexts. Flow is intentionally designed as a supervisory and structuring tool that supports clinical reasoning without generating autonomous diagnoses or treatment decisions. Core design principles include transparency of AI outputs, clinician control, traceability of information, and clearly defined boundaries of clinical responsibility. Ethical considerations related to data governance, explainability, bias mitigation, and human–AI collaboration are embedded at both technical and organizational levels.

The presentation synthesizes key insights emerging from clinical field experience regarding clinician expectations, perceived risks, and conditions for acceptability of AI tools. Particular attention is given to tensions between efficiency and clinical nuance, automation and professional judgment, and innovation and regulatory responsibility. These insights highlight the importance of framing AI as cognitive support rather than as a decision-making authority.

In conclusion, this work argues that the future of AI in neuropsychological and mental health assessment depends less on algorithmic sophistication than on ethical design, clinician trust, and alignment with professional values. Future directions include empirical evaluation of acceptability and long-term human–AI collaboration in clinical practice.

Keywords

Artificial intelligence; Neuropsychology; Mental health assessment; Clinical reasoning; Ethics; Human-centered AI; Clinician acceptability; Decision support