Abstract 069: AI‐driven LVO Detection: Assessment of Clinical Utility in Diverse Prevalence Settings
In recent years, several AI‐based algorithms for detecting intracranial large vessel occlusion (LVO) have been developed, aiming to accelerate the triage of LVO patients and regarded as promising tools to detect LVO in settings where trained stroke specialists are unavailable. LVO‐detection algorithms require high specificity to increase positive predictive value (PPV). However, predictive values are influenced by the prevalence of the condition: if in the absence of trained clinicians, imaging is used to screen a broader population of suspected stroke patients, algorithm performance will diminish.
Authors
- Jorge Mayol (ORCID: https://orcid.org/0000-0002-7739-287X)
- Federica Rizzo (ORCID: https://orcid.org/0000-0002-5945-8222)
- Marta Rubiera (ORCID: https://orcid.org/0000-0001-8100-9477)
- Álvaro García‐Tornel (ORCID: https://orcid.org/0000-0003-3633-3002)
- R. Simonetti
- Francesco Diana (ORCID: https://orcid.org/0000-0002-3245-917X)
- María Jesús Rodrigo (ORCID: https://orcid.org/0000-0002-4009-3075)
- Marc Ribó (ORCID: https://orcid.org/0000-0001-9242-043X)
- Miguel Requena (ORCID: https://orcid.org/0000-0002-4490-6029)
Institutions
- Vall d'Hebron Hospital Universitari (ES)
Publication Details
- Journal
- Stroke Vascular and Interventional Neurology
- Published
- 2024-11-01
- DOI
- https://doi.org/10.1161/svin.04.suppl_1.069
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 0.00