Review appraises CT-based prediction of hematoma expansion and outcomes after spontaneous intracerebral hemorrhage
This review compares contrast-enhanced and non-contrast CT markers, composite models and artificial intelligence approaches for predicting hematoma expansion and adverse outcomes after spontaneous intracerebral hemorrhage, and discusses aging-related factors; it reports no new patient data. [1]
Automatic summary · not editorially reviewed
The key findings
The review compares contrast-enhanced CT markers (spot sign, leakage sign, iodine sign) with non-contrast CT markers for predicting hematoma expansion after spontaneous intracerebral hemorrhage. [1]
The review assesses composite prediction models and artificial intelligence approaches, emphasizing limitations such as small cohorts, overfitting and insufficient external validation. [1]
The authors discuss how vascular aging, cerebral amyloid angiopathy, frailty, anticoagulant exposure and age-associated vulnerability may influence hematoma expansion risk and outcome prediction. [1]
The source at a glance
- Study type / format
- Narrative review (abstract only)
- Population / context
- People with spontaneous intracerebral hemorrhage, as addressed in a review of CT-based prediction.
- Summary based on
- Study abstract
- Original publication
- August 15, 2026
CT markers compared for hematoma expansion risk
The review evaluates CT-based prediction of hematoma expansion and adverse outcomes after spontaneous intracerebral hemorrhage. It compares contrast-enhanced markers, namely the spot sign, leakage sign and iodine sign, with non-contrast CT markers such as the blend sign, black hole sign, island sign, satellite sign, hypodensity sign, swirl sign, hematoma shape and density heterogeneity. The comparison focuses on sensitivity, specificity, reproducibility, availability and clinical applicability. The authors note that current evidence is dispersed across conventional CT signs, composite scores, radiomics, machine learning and deep learning approaches, many reported descriptively without sufficient comparison of clinical utility, validation quality or translational readiness.
Supporting source [1]
Composite models and AI: validation and workflow limits
The review assesses composite prediction models and artificial intelligence approaches. It emphasizes limitations related to small cohorts, overfitting, dataset heterogeneity, insufficient external validation, interpretability and workflow integration. Because the scope is GeroScience, the authors also discuss how vascular aging, cerebral amyloid angiopathy, frailty, anticoagulant exposure and age-associated vulnerability to secondary injury may influence hematoma expansion risk and outcome prediction. They state that future progress will require interpretable, multimodal, prospectively validated models integrating CT imaging, clinical variables, biomarkers and aging-related factors to support individualized management of intracerebral hemorrhage.
Supporting source [1]
Abstract-only appraisal without new patient data
This is a review that critically evaluates CT-based prediction of hematoma expansion and adverse outcomes after spontaneous intracerebral hemorrhage. The abstract describes the compared marker types, models and discussed aging-related factors, but reports no cohort size, follow-up duration, effect estimates or numerical results. Only the abstract was reviewed; the full text was not reviewed. The abstract does not establish that any marker or model improves patient outcomes, and the discussed aging-related influences are presented as possible factors, not as demonstrated causes.
Supporting source [1]
How to interpret this
Only the abstract was reviewed; the full text was not reviewed. The abstract reports no cohort size, follow-up, effect estimates or numerical results, and it does not show that any marker or model improves outcomes. Discussed aging-related influences are presented as possible factors, not demonstrated causes.
Based on the study abstract; the full paper, tables and supplementary data were not reviewed.
Source & supporting passages
Shuai Liu, Weihao Fu, Xiaoying Tang, Yingwei Fan · doi.org · August 15, 2026