An Explainable Agentic AI Framework for Uncertainty-Aware and Abstention-Enabled Acute Ischemic Stroke Imaging Decisions

An Explainable Agentic AI Framework for Uncertainty-Aware and Abstention-Enabled Acute Ischemic Stroke Imaging Decisions 1st Md Rashadul Islam Department of Computer Science and Engineering Daffodil International University islam15-6062@s.diu.edu.bd Dhaka, Bangladesh Abstract—Artificial intelligence (AI) models have demonstrated considerable potential in the imaging of acute ischemic stroke, especially in the detection and segmentation of lesions via computed tomography (CT) and magnetic resonance imaging (MRI). Never- theless, the majority of existing approaches operate as black-box predictors, providing deterministic outputs without transparency regarding predictive uncertainty or the establishment of explicit protocols for decision rejection when predictions are ambiguous. This de- ficiency presents considerable safety and trust issues within the context of high-stakes emergency radiology, where inaccuracies in automated decision-making could conceivably lead to negative consequences in clinical settings. [1], [2]. In this paper, we introduce an explainable agen- tic AI framework targeted at uncertainty-aware and abstention-based decision-making in AIS imaging. It is based on a multistage agentic pipeline. In this framework, a perception agent performs lesion-aware image analysis, an uncertainty estima- tion agent estimates the predictive confidence at the slice level and a decision agent dynamically decides whether to make or withhold the prediction based on prescribed uncertainty thresholds. This approach is different from previous stroke imaging frameworks, which have primarily aimed to improve the accuracy of segmentation or classification. [3], [4], Our framework explicitly emphasizes clinical safety, transparency, and decision-making processes that are congruent with hu- man values. We validate the practicality and interpretability of our framework through qualitative and case-based ex- aminations of typical stroke imaging scenarios. This examination demonstrates a natural correlation be- tween uncertainty-driven abstention and the existence of lesions, fluctuations in image quality, and the specific anatomical definition being analyzed. Furthermore, the system integrates an explanation mode, offering visual and structural justifications to bolster decision-making, thereby addressing a crucial limitation observed in existing uncertainty-aware medical imaging systems: the absence of actionable interpretability. [5], [6]. This research does not claim to establish a high- performance benchmark; instead, it presents agentic control, uncertainty-awareness, and selective absten- tion as essential design principles for the creation of safe and reliable MI-AI. Our results support the idea that incorporating explicit stalling behavior within agentic architectures could accelerate the development of clinically deployable AI systems for acute stroke intervention. Keywords: Acute Ischemic Stroke; Medical Imaging AI; Agentic Artificial Intelligence; Uncertainty Estimation; Abstention Mechanisms; Explainable AI; Clinical Decision Support; Safety-Critical AI I. Introduction Acute ischemic stroke continues to be one of the leading causes of mortality and long-term disability worldwide, placing a significant burden on patients, caregivers, and healthcare systems. Accurate and timely interpretation of medical imaging, particularly computed tomography (CT), CT angiography (CTA), and magnetic resonance imaging (MRI).Artificial intelligence (AI) has emerged as a promising tool to support radiological assessment by enabling automated lesion detection, segmentation, and triage from stroke imaging data [3], [4]. While deep learning has advanced significantly in the domain of medical imaging, the majority of current AI systems for stroke diagnosis function as deterministic pre- dictors, yielding consistent results without accounting for variables such as image quality, lesion ambiguity, or shifts in data distribution. This approach contrasts sharply with the practical realm of clinical radiology, where experienced radiologists often delay decision-making until additional imaging is obtained or process complex cases through multiple stages to accommodate substantial diagnostic uncertainty. In the absence of explicit mechanisms to iden- tify images with uncertainty and abstain from decision- making, issues of safety and trust become increasingly critical, particularly in emergency situations where erro- neous automated decisions could have profound clinical consequences. [1], [2]. Recent work has emphasized the importance of uncer- tainty quantification in medical AI and the necessity of distinguishing reliable from non-reliable predictions [7]– [9]. In medical imaging, uncertainty-aware approaches arXiv:2601.01008v1 [eess.IV] 3 Jan 2026 such as Bayesian neural networks and deep ensembles (DE) have been studied. However, they usually assume that the results can only be expressed as a numerical confidence level and would no

This content is AI-processed based on open access ArXiv data.