|
Reinforcement Learning-Based Clinical Decision Support for Real-Time, Context-Aware Treatment Suggestions
|
|
| Author:
|
, V.RAVIKUMAR, P. BALAKUMAR, R. DHANALAKSHMI
|
| Abstract:
|
Clinical decision support systems (CDSS) provide a critical role in improving medical decision making, but the traditional rule-based approach is not adaptable to changing patient conditions. To address these issues, this paper presents a Reinforcement Learning-Based Clinical Decision Support System (RL-CDSS) offering real-time, context-aware treatment suggestion. Multi agent reinforcement learning (MARL) is integrated to solve the system under this framework, with contextual multi armed bandits (CMAB) used for its adaptive decision making. Patient monitoring is improved with IoT-driven real-time sensor data, and both can be done with a federated learning approach in which the model is trained with respect for privacy across multiple healthcare institutions. Additionally, an Explainable AI (XAI) framework helps make physician trust by providing interpretable treatment recommendations. Finally, the proposed system is validated against real-world clinical data, and deployed on a commercial system where traditional CDSS falls. This research points out the possibility of using the RL assisted CDSS to transform personalized healthcare, medical workflows, and consequently patient outcomes, and surmounts the obstacles of data privacy, real-time adaptability, as well as physician acceptance.
|
| Keyword:
|
CMAB, CDSS, Reinforcement, Real-Time, Context-Aware, XAI
|
| EOI:
|
-
|
| DOI:
|
https://doi.org/10.31838/ijpr/2019.11.01.309
|
| Download:
|
Request For Article
|
|
|