Dynamic Risk Stratification for Short-Term Deterioration After Acute Stroke: Incremental Value of Radiology-Report-Derived Neuroimaging Phenotypes Beyond Dynamic ICU States, a retrospective cohort study with temporal validation
Abstract
BackgroundPrediction models after stroke increasingly add neuroimaging information to bedside data. Whether findings documented in radiology reports improve short-term risk stratification beyond a patients evolving physiological state is unknown.
MethodsWe studied 4,827 adults with acute ischaemic stroke, intracerebral haemorrhage or subarachnoid haemorrhage in MIMIC-IV (2008-2019). At eight landmarks from 6 to 48 hours after ICU admission, models ranked patients by their risk of deterioration or death within 24 hours. The baseline model combined demographic and clinical characteristics, a previously validated dynamic ICU state representation, and the physiology, treatment and imaging status at each landmark. A locally run language model extracted imaging phenotypes from 6,992 head CT and MRI reports; three phenotypes and four location categories met a pre-specified reliability criterion for automated extraction in independent validation and entered the model. The primary outcome was the difference in the proportion of events captured among the 10% of patients ranked highest at each landmark, in temporal validation (2017-2019; 8,923 patient-landmarks, 666 events).
ResultsAmong the highest-ranked 10%, adding imaging phenotypes captured no additional event (348 of 666 with and without them): a difference of 0.00 percentage points (95% CI -2.20 to +2.01). Discrimination (AUROC 0.855 vs 0.849), calibration and net benefit were similar. No prespecified sensitivity analysis, stroke subtype or availability-adjusted analysis showed statistically supported incremental capture. In a post hoc redundancy analysis, imaging phenotypes improved capture when added to demographic and clinical characteristics alone (+12.91 points, 7.54 to 17.81), but this gain was no longer evident once the dynamic state representation was included.
ConclusionsNeuroimaging phenotypes derived from routine radiology reports were associated with short-term deterioration, but we found no evidence that they added risk stratification value beyond the dynamic ICU state and contemporaneous physiology.
Related
- Preoperative risk stratification for postoperative inflammatory–nutritional deterioration after colorectal cancer surgery: an interpretable machine learning study
- Incremental Value of Radiation Dose for Prognostic Stratification Beyond Performance Status in Palliative Radiation Therapy for Bone Metastases
- Neuroimaging markers associated with early neurological deterioration in acute isolated pontine infarction: a systematic review and meta-analysis
- Clinical Phenotypes, Neuroimaging, and Outcomes in Prolonged and Persistent Non-Motor Migraine Aura: A Systematic Review and Episode-Level Synthesis
- Endocannabinoid System Genes in Dementia: A Systematic Mendelian Randomization Study of Functional Variants and Tissue-Specific Expression Across Clinical, Molecular, and Neuroimaging Phenotypes
BibTeX
@misc{p2026dynamic,
title = {{Dynamic Risk Stratification for Short-Term Deterioration After Acute Stroke: Incremental Value of Radiology-Report-Derived Neuroimaging Phenotypes Beyond Dynamic ICU States, a retrospective cohort study with temporal validation}},
author = {P. LEI and Y. XU and Y. ZHANG and N. RAY},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.08.26365059},
url = {https://www.medrxiv.org/content/10.64898/2026.10.08.26365059v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Dynamic Risk Stratification for Short-Term Deterioration After Acute Stroke: Incremental Value of Radiology-Report-Derived Neuroimaging Phenotypes Beyond Dynamic ICU States, a retrospective cohort study with temporal validation
AU - P. LEI
AU - Y. XU
AU - Y. ZHANG
AU - N. RAY
PY - 2026
DA - 2026/10/08
PB - medRxiv
DO - 10.64898/2026.10.08.26365059
UR - https://www.medrxiv.org/content/10.64898/2026.10.08.26365059v1
AB - BackgroundPrediction models after stroke increasingly add neuroimaging information to bedside data. Whether findings documented in radiology reports improve short-term risk stratification beyond a patients evolving physiological state is unknown. MethodsWe studied 4,827 adults with acute ischaemic stroke, intracerebral haemorrhage or subarachnoid haemorrhage in MIMIC-IV (2008-2019). At eight landmarks from 6 to 48 hours after ICU admission, models ranked patients by their risk of deterioration or death within 24 hours. The baseline model combined demographic and clinical characteristics, a previously validated dynamic ICU state representation, and the physiology, treatment and imaging status at each landmark. A locally run language model extracted imaging phenotypes from 6,992 head CT and MRI reports; three phenotypes and four location categories met a pre-specified reliability criterion for automated extraction in independent validation and entered the model. The primary outcome was the difference in the proportion of events captured among the 10% of patients ranked highest at each landmark, in temporal validation (2017-2019; 8,923 patient-landmarks, 666 events). ResultsAmong the highest-ranked 10%, adding imaging phenotypes captured no additional event (348 of 666 with and without them): a difference of 0.00 percentage points (95% CI -2.20 to +2.01). Discrimination (AUROC 0.855 vs 0.849), calibration and net benefit were similar. No prespecified sensitivity analysis, stroke subtype or availability-adjusted analysis showed statistically supported incremental capture. In a post hoc redundancy analysis, imaging phenotypes improved capture when added to demographic and clinical characteristics alone (+12.91 points, 7.54 to 17.81), but this gain was no longer evident once the dynamic state representation was included. ConclusionsNeuroimaging phenotypes derived from routine radiology reports were associated with short-term deterioration, but we found no evidence that they added risk stratification value beyond the dynamic ICU state and contemporaneous physiology.
ER -