Both Infarcted and Noninfarcted Brain Regions Contribute to Deep Learning–Based MRI Prediction of Acute Stroke Outcome
While the best performance came from by using the full brain imaging volume, we demonstrate that the infarct location, its signal characteristics, and importantly, the noninfarcted regions all contribute to the predictions. The noninfarcted areas may be a proxy for overall brain health and resilience, containing important information about potential outcomes.
Authors
- Maarten G. Lansberg (ORCID: https://orcid.org/0000-0002-3545-6927)
- Jeremy Josef Heit (ORCID: https://orcid.org/0000-0003-1055-8000)
- Yongkai Liu (ORCID: https://orcid.org/0000-0003-4478-4006)
- Yannan Yu (ORCID: https://orcid.org/0000-0002-1718-6505)
- David S. Liebeskind (ORCID: https://orcid.org/0000-0002-5109-8736)
- Gregory W. Albers (ORCID: https://orcid.org/0000-0003-0263-4632)
- Michael E. Moseley (ORCID: https://orcid.org/0000-0002-9163-8741)
- Henk van Voorst (ORCID: https://orcid.org/0000-0002-2647-3557)
- Max Wintermark (ORCID: https://orcid.org/0000-0002-6726-3951)
- Greg Zaharchuk (ORCID: https://orcid.org/0000-0001-5781-8848)
- Z B Zhang (ORCID: https://orcid.org/0009-0002-4427-3847)
- Shuo Li (ORCID: https://orcid.org/0000-0003-1152-5593)
- Haotian Feng (ORCID: https://orcid.org/0000-0001-8133-4114)
- Bin Jiang (ORCID: https://orcid.org/0000-0002-8507-0459)
- S. Luo (ORCID: https://orcid.org/0000-0001-9735-4670)
Institutions
- The University of Texas MD Anderson Cancer Center (US)
- University of California, Los Angeles (US)
- University of California, San Francisco (US)
- UCLA Health (US)
- Case Western Reserve University (US)
- Stanford University (US)
Publication Details
- Journal
- American Journal of Neuroradiology
- Published
- 2025-11-06
- DOI
- https://doi.org/10.3174/ajnr.a8896
- Primary Topic
- Acute Ischemic Stroke Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00