CardAIc-Agents: A Multimodal Framework with Hierarchical Adaptation for Cardiac Care Support

2025-08-18

AI Summary

Cardiovascular diseases impose a substantial global health burden, yet early intervention is hindered by healthcare resource shortages and the limited clinical adaptability of existing AI agents. To address this, we propose a multimodal adaptive intelligent agent framework tailored for cardiac diagnosis and treatment. Methodologically, it features: (1) a hierarchical dynamic decision-making architecture enabling personalized pathway planning based on real-time detection outputs; (2) integrated CardiacRAG—a retrieval-augmented knowledge base—complemented by multidisciplinary discussion tools and an interactive visualization review panel to support clinical-feedback-driven continual optimization; and (3) synergistic fusion of vision-language models, external tool invocation, multi-agent collaboration, and on-demand multimodal generation—including medical image synthesis. Extensive experiments across three clinical datasets demonstrate significant improvements in task execution efficiency and clinical alignment, outperforming state-of-the-art vision-language models, advanced agent systems, and fine-tuned baseline models.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Cardiovascular diseases (CVDs) remain the foremost cause of mortality worldwide, a burden worsened by a severe deficit of healthcare workers. Artificial intelligence (AI) agents have shown potential to alleviate this gap via automated early detection and proactive screening, yet their clinical application remains limited by: 1) prompt-based clinical role assignment that relies on intrinsic model capabilities without domain-specific tool support; or 2) rigid sequential workflows, whereas clinical care often requires adaptive reasoning that orders specific tests and, based on their results, guides personalised next steps; 3) general and static knowledge bases without continuous learning capability; and 4) fixed unimodal or bimodal inputs and lack of on-demand visual outputs when further clarification is needed. In response, a multimodal framework, CardAIc-Agents, was proposed to augment models with external tools and adaptively support diverse cardiac tasks. Specifically, a CardiacRAG agent generated general plans from updatable cardiac knowledge, while the chief agent integrated tools to autonomously execute these plans and deliver decisions. To enable adaptive and case-specific customization, a stepwise update strategy was proposed to dynamically refine plans based on preceding execution results, once the task was assessed as complex. In addition, a multidisciplinary discussion tool was introduced to interpret challenging cases, thereby supporting further adaptation. When clinicians raised concerns, visual review panels were provided to assist final validation. Experiments across three datasets showed the efficiency of CardAIc-Agents compared to mainstream Vision-Language Models (VLMs), state-of-the-art agentic systems, and fine-tuned VLMs.

Problem

• Addressing healthcare worker shortage via AI for cardiac care
• Overcoming limitations in clinical AI role assignment and workflows
• Enhancing adaptive reasoning and multimodal capabilities in diagnosis

Innovation

• Multimodal framework integrating external tools
• Hierarchical agents with adaptive stepwise updates
• Dynamic visual review panels for validation