From Sensors to Agency: An AI-Driven Ecosystem for Predictive and Personalized Management of Neurodegenerative Disorders

The increasing prevalence of neurodegenerative disorders such as Alzheimer’s Disease (AD) and Parkinson’s Disease (PD) demands innovative technological and managerial responses that extend beyond clinical diagnosis. Recent advances in ML and sensor-based monitoring have enabled early detection and tracking of disease progression through continuous physiological and behavioural data streams (Esteva et al., 2021). Yet, most of these systems remain passive regarding the capability of analyzing data but not acting autonomously. The emerging field of Agentic AI allows agents to make adaptive, goal-directed decisions, introducing a new opportunity. For example, creating healthcare technologies that can learn, reason, and intervene dynamically to support patients and healthcare systems (Hughes et al., 2025

Karunanayake, 2025).

ML applications in neurodegenerative disease detection have demonstrated high diagnostic accuracy (Rajkomar et al., 2019

Esteva et al., 2021). They are often isolated from real-world healthcare operations and lack continuous decision feedback loops. Current sensor networks, though powerful, suffer from data fragmentation and limited integration across multiple care environments. Moreover, the engineering of adaptive cyber-physical systems capable of coordinating data flow, resource use, and patient feedback in real time remains underexplored (Li et al., 2025). These gaps highlight the need for a technological framework that merges the strengths of agentic intelligence, sensor fusion, and system-level orchestration.

This research proposes the design of an agentic AI-driven neurocare ecosystem that operates across three layers: sensing, cognition, and coordination. The sensing layer integrates multimodal signals from wearable and ambient sensors, capturing physiological and behavioural markers relevant to PD or AD. The cognition layer employs reinforcement learning and multimodal transformers to learn patient-specific patterns, detect anomalies, and predict clinical events. The coordination layer functions as an intelligent orchestration hub dynamically allocating computational and clinical resources, prioritizing interventions, and synchronizing updates across medical teams and digital interfaces. This layered structure ensures engineering robustness (through system design and control optimization) and informational coherence (through data harmonization and feedback). It enhances system autonomy, resilience, and adaptability in continuous healthcare monitoring. It introduces mechanisms for data-driven coordination and resource optimization. This research aims to redefine how agentic AI technologies can be engineered and governed to manage chronic neurological conditions for transforming reactive care models into proactive, learning-driven ecosystems.

References

Esteva, A., Chou, K., Yeung, S., Naik, N., Madani, A., Mottaghi, A., Liu, Y., Topol, E., Dean, J. and Socher, R. (2021). Deep learning-enabled medical computer vision. NPJ digital medicine, 4(1), 5.

Hughes, L., Dwivedi, Y. K., Malik, T., Shawosh, M., Albashrawi, M. A., Jeon, I., ... & Walton, P. (2025). AI agents and agentic systems: A multi-expert analysis. Journal of Computer Information Systems, 1-29.

Karunanayake, N. (2025). Next-generation agentic AI for transforming healthcare. Informatics and Health, 2(2), 73-83.

Li, A., Lian, J., & Vardhanabhuti, V. (2025). Multi-modal machine learning approach for early detection of neurodegenerative diseases leveraging brain MRI and wearable sensor data. PLOS Digital Health, 4(4), e0000795.

Rajkomar, A., Dean, J., & Kohane, I. (2019). Machine learning in medicine. New England Journal of Medicine, 380(14), 1347-1358.