Artificial Intelligence in Acute Stroke: Applications and Implications for the Portuguese National Healthcare System
- 1Faculdade de Medicina. Universidade de Lisboa. Lisbon.
- 2Department of Neuroradiology. Hospital Santa Maria. Unidade Local de Saúde de Santa Maria. Lisbon.
Abstract
Introduction: Acute stroke is a leading cause of death and long-term disability in Portugal. Rapid and accurate imaging is critical to guide acute stroke management, and artificial intelligence has shown growing potential to support triage, quantification, and therapeutic decision-making. This study aimed to characterize the current use of artificial intelligence in acute stroke imaging workflow in Portuguese public hospital neuroradiology departments, and to identify perceived benefits, implementation barriers and prospects. Methods: A descriptive, cross-sectional study was conducted using an online questionnaire distributed between January and April 2026 to neuroradiology departments of Portuguese public hospitals with stroke intervention units. The questionnaire covered department characterization, current use of artificial intelligence, and perceptions regarding future applications. Responses were analyzed descriptively. Results: Ten hospitals from the North, Centre, Lisbon Metropolitan Area, South and insular regions were included. All operated under a stroke code protocol, used computed tomography as the first-line modality, and provided 24-hour neuroradiology coverage through a combined on-site and teleradiology model. Only two hospitals used dedicated artificial intelligence software, and one was in an implementation phase. Three additional centers used computational tools with artificial intelligence functionalities without identifying them as such. Prior contact with artificial intelligence was reported by 80% of respondents, and all anticipated its integration into routine practice. The most frequently cited applications were automated ASPECTS, core/penumbra quantification, large vessel occlusion identification and hemorrhage detection. Faster time-to-treatment and support for teleradiology workflows were the most perceived benefits. Opinions on funding were divided between the hospital institution and the state. Formal staff training was reported only by one center. Conclusion: Despite high awareness and a well-established organizational framework, adoption of artificial intelligence remains limited. Coordinated procurement strategies, structured training programs and shared operational definitions of artificial intelligence are needed to enable broader and more equitable integration into stroke care.
Related
- Impact of Artificial Intelligence on Operational Efficiency and Cost Reduction in U.S. Healthcare Organizations
- Bias Mitigation Across the Healthcare Artificial Intelligence Lifecycle: A Structured Narrative Review
- Predictors of Intentions to Learn and Adopt Medical Artificial Intelligence Among Healthcare Professionals in Ghana: A Descriptive Cross‐Sectional Study
- Artificial intelligence-enabled mental health applications for adolescents: a scoping review of safety, effectiveness and clinical implementation
- Artificial Intelligence Applications in Laser-Assisted Refractive Surgery—A Comprehensive Review
BibTeX
@article{ribeiro2026artificial,
title = {{Artificial Intelligence in Acute Stroke: Applications and Implications for the Portuguese National Healthcare System}},
author = {Laura Ribeiro and João Madureira and Sofia Reimão},
journal = {Acta Médica Portuguesa},
year = {2026},
month = nov,
doi = {10.20344/amp.24990},
url = {https://doi.org/10.20344/amp.24990},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - Artificial Intelligence in Acute Stroke: Applications and Implications for the Portuguese National Healthcare System
AU - Laura Ribeiro
AU - João Madureira
AU - Sofia Reimão
PY - 2026
DA - 2026/11/02
JO - Acta Médica Portuguesa
DO - 10.20344/amp.24990
UR - https://doi.org/10.20344/amp.24990
AB - Introduction: Acute stroke is a leading cause of death and long-term disability in Portugal. Rapid and accurate imaging is critical to guide acute stroke management, and artificial intelligence has shown growing potential to support triage, quantification, and therapeutic decision-making. This study aimed to characterize the current use of artificial intelligence in acute stroke imaging workflow in Portuguese public hospital neuroradiology departments, and to identify perceived benefits, implementation barriers and prospects. Methods: A descriptive, cross-sectional study was conducted using an online questionnaire distributed between January and April 2026 to neuroradiology departments of Portuguese public hospitals with stroke intervention units. The questionnaire covered department characterization, current use of artificial intelligence, and perceptions regarding future applications. Responses were analyzed descriptively. Results: Ten hospitals from the North, Centre, Lisbon Metropolitan Area, South and insular regions were included. All operated under a stroke code protocol, used computed tomography as the first-line modality, and provided 24-hour neuroradiology coverage through a combined on-site and teleradiology model. Only two hospitals used dedicated artificial intelligence software, and one was in an implementation phase. Three additional centers used computational tools with artificial intelligence functionalities without identifying them as such. Prior contact with artificial intelligence was reported by 80% of respondents, and all anticipated its integration into routine practice. The most frequently cited applications were automated ASPECTS, core/penumbra quantification, large vessel occlusion identification and hemorrhage detection. Faster time-to-treatment and support for teleradiology workflows were the most perceived benefits. Opinions on funding were divided between the hospital institution and the state. Formal staff training was reported only by one center. Conclusion: Despite high awareness and a well-established organizational framework, adoption of artificial intelligence remains limited. Coordinated procurement strategies, structured training programs and shared operational definitions of artificial intelligence are needed to enable broader and more equitable integration into stroke care.
SN - 0870-399X
SN - 1646-0758
ER -