Posterior circulation stroke: machine learning-based detection of early ischemic changes in acute non-contrast CT scans
Problem
Visual detection of early ischemic changes in posterior circulation stroke on non-contrast CT is unreliable due to reader variability, artifacts, and low sensitivity.
Why it matters
Basilar artery occlusion is deadly and missed early signs delay triage, therapy planning, and outcome prediction.
What the AI did
A random forest model trained on 1218 quantitative CT image features detected early ischemic changes in pc-ASPECTS regions, evaluated with fivefold cross-validation on 552 regions from 69 patients.
Results
not stated in source (abstract truncated before performance figures)
Impact
Retrospective validation on 69 patients; clinical deployment or external validation not stated.
Record
| Venue | Journal of Neurology |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Helge C. Kniep, Peter Bernhard Sporns, Gabriel Broocks, André Kemmling, Jawed Nawabi, Thilo Rusche, Jens Fiehler, Uta Meta Paula Hanning |
| Citations | 36 (OpenAlex, at import time) |
| References | 31 |
| Model or method | Random forest algorithms |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1007/s00415-020-09859-4 |
| Open access | hybrid |
| Keywords | stroke, non-contrast CT, random forest, pc-ASPECTS, machine learning |
Importance rating
significant — thousands to millions affected
Aids triage and therapy planning for basilar artery occlusion patients where visual CT reading is unreliable; clinical benefit clear but abstract gives no scale or mortality figures.