{
  "abstract": "Introduction Artificial Intelligence (AI) is transforming interventional neuroradiology (INR) by enhancing stroke triage, aneurysm detection, procedural planning, and patient management.Aim of Study While AI has demonstrated significant potential, challenges remain in clinical integration, data standardization, and regulatory approval.Method A review of recent literature and AI-based applications in INR was conducted, focusing on diagnostic accuracy, workflow efficiency, and real-time decision support. Insights from neuroradiologists were analyzed to identify current barriers and future needs.Results AI has shown high sensitivity and specificity in detecting ischemic strokes, aneurysms, and arteriovenous malformations (AVMs). Machine learning models assist in predicting thrombus composition, guiding thrombectomy techniques, and automating imaging post-processing, reducing interpretation time and improving outcomes. However, data heterogeneity, clinician acceptance, and regulatory constraints limit widespread adoption.Conclusion Future AI applications include robotic-assisted interventions, real-time AI-driven guidance, and predictive analytics for patient outcomes. Standardizing datasets and leveraging federated learning could improve model reliability. A multidisciplinary approach involving AI developers, clinicians, and regulators is essential to ensure safe and interpretable AI integration. AI has the potential to revolutionize INR by improving efficiency, precision, and patient outcomes. Addressing technical, ethical, and clinical challenges will be key to its successful implementation, ensuring AI supports rather than replaces neuroradiologists.Conflict of Interest Yes I am a board member and CEO of RadioLogs",
  "authors": [
    {
      "affiliations": [
        "Erasme Hospital, Brussels, Belgium"
      ],
      "name": "Adrien Guenego"
    }
  ],
  "title": "A17 Future directions of AI in interventional neuroradiology (INR), input from a young INR/startuper",
  "uid": "a73e36ee-4e2a-5c93-a444-11d65d3d9c78"
}