Agentic Explainable AI Framework for Intelligent Clinical Decision Support Using Large Language Models and Multimodal Healthcare Data
Journal of Intelligent Decision Making and Information Science · 15.07.2026 · 3 (2) · 1887-1897
DOI: 10.17762/jidmis.v3.2454 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Intelligent Decision Making and Information Science, online erschienen am 15.07.2026. Verfasst von D. Saisanthiya. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The swift expansion of multimodal health data, such as EHRs, medical images, lab reports, physiological signals, and clinical notes, has opened up many avenues for intelligent clinical decision support systems. However, the traditional AI models are generally considered to be black-box systems, which does not equate to clinician’s confidence and ease of use in real-world health care situations. Therefore, this study presents an Agentic Explainable Artificial Intelligence (AXAI) Framework for intelligent clinical decision support by integrating Large Language Models (LLMs), multimodal healthcare data, and autonomous AI agents. The presented framework uses specialized agents intended for data collection, data processing, multimodal feature extraction, explainability generation, and validation of decisions. Structured clinical information is combined with unstructured texts and imaging information to obtain a detailed description the patients. Multimodal fusion is used to combine structured clinical data with unstructured medical text and images. This is done to ensure a full representation of the patient. The reasoning agents use LLMs to give diagnostic suggestions and treatment recommendations based on the available clinical guidelines. Explainable AI methods to provide feature attribution, reasoning based on facts, measurement of confidence, and natural language explanations backing recommendations have been added to the framework to increase reliability and trust of clinicians. In addition, there is an autonomous validation entity responsible for assessing consistency of the decisions-making process, identifying contradictory evidence, and allowing human verification of the system before making recommendations. Experimental tests show that the suggested framework produces better diagnostic accuracy, interpretability, robustness, and decision consistency than traditional clinical decision systems based on machine learning and deep learning. The agentic design also allows for more adaptation to different healthcare scenarios while cutting down on diagnostic uncertainty through the cooperative reasoning of different intelligent agents. The proposed framework introduces a reliable, scalable, and transparent method for innovative intelligent healthcare systems by employing multimodal learning, explainable AI, and autonomous decision-making with LLM to aid healthcare professionals in providing effective, personalized healthcare for patients.
Zitieren
APA
Saisanthiya, D. (2026). Agentic Explainable AI Framework for Intelligent Clinical Decision Support Using Large Language Models and Multimodal Healthcare Data. Journal of Intelligent Decision Making and Information Science, 3(2), 1887-1897. https://doi.org/10.17762/jidmis.v3.2454
Vancouver
Saisanthiya D. Agentic Explainable AI Framework for Intelligent Clinical Decision Support Using Large Language Models and Multimodal Healthcare Data. Journal of Intelligent Decision Making and Information Science. 2026;3(2):1887-1897. doi:10.17762/jidmis.v3.2454
BibTeX
@article{Saisanthiya2026,
author = {Saisanthiya, D.},
title = {Agentic Explainable AI Framework for Intelligent Clinical Decision Support Using Large Language Models and Multimodal Healthcare Data},
journal = {Journal of Intelligent Decision Making and Information Science},
year = {2026},
volume = {3},
number = {2},
pages = {1887--1897},
doi = {10.17762/jidmis.v3.2454}
}
EUVIMED