WSO - February 2024 - 227

Song et al.
227
were independent predictors of HE (odds ratio (OR) = 0.98, 95% confidence interval (CI) = 0.97-0.99; OR = 1.73,
95% CI = 1.28-2.35; OR = 0.87, 95% CI = 0.83-0.92; OR = 0.42, 95% CI = 0.28-0.62; OR = 7.82, 95% CI = 5.86-10.42,
respectively). The optimal cut-off point for FIM in predicting HE was 0.63, with sensitivity, specificity, PPV, NPV, and
AUC values of 0.69, 0.89, 0.71, 0.88, and 0.83, respectively.
Conclusion: The FIM adjusted for time since symptom onset is a significant predictor of HE. Its use may allow improved
prediction of those patients with ICH who develop HE, and the score may be clinically applicable in the management of
patients with ICH.
Keywords
Intracerebral hemorrhage, hematoma expansion, computed tomography, marker
Received: 11 July 2023; accepted: 4 September 2023
Introduction
Hematoma expansion (HE) is seen in one-third of patients
with intracerebral hemorrhage (ICH) and is directly linked
to disability.1-3 Biomarkers observed on non-contrast computed
tomography (CT) have been an important component
of clinical protocols for HE assessment in recent years.2,4,5
Despite its appealing proposals in clinical settings, the
overall performance of imaging markers is unsatisfactory.4
Meanwhile, studies have reported that imaging markers can
be used in combination to improve the prediction of HE;
however, their performance was poor.6,7
The time from onset to initial CT scan (onset-to-CT time
(OCT)) is another important consideration of HE;8 the
shorter the OCT, the higher the risk of HE.9 Furthermore,
studies showed that the predictive accuracy of imaging
markers for HE was also affected by OCT,10,11 suggesting
that only imaging markers or OCT may not be sufficient for
HE prediction. Therefore, a new indicator that simultaneously
considers imaging markers and OCT for the prediction
of HE warrants further exploration. We hypothesized
that the frequency of imaging markers (FIM), that is, the
ratio of the number of imaging markers and OCT, is a novel
predictor of HE.
Aims
This study aimed to determine the impact of FIM on HE
prediction and assess its performance.
Methods
In this multicenter retrospective cohort study, consecutive
patients with spontaneous ICH were included from January
2018 to August 2022. The six hospitals were located in the
provinces of Hubei (Huangshi Central Hospital, Xiangyang
Central Hospital, and Xiangyang No. 1 People's Hospital),
Anhui (The First Affiliated Hospital and the Second
Affiliated Hospital of Anhui Medical University), and
Hebei (Tangshan Gongren Hospital), respectively. The
Institutional Review Board at the local ethics committees
granted an exemption to obtain informed consent due to the
use of deidentified data (Ethics Approval ID: 2022-22,
2021-036, XYYYE20220081, PJ2022-09-30, YX2023134,
and GRYY-LL-KJ2022-K820, respectively).
Demographic characteristics (age, sex), prior history
(alcohol, smoking, diabetes mellitus), laboratory results on
admission (glucose level, platelet count, and international
normalized ratio), and clinical features on admission (systolic
and diastolic blood pressure values and baseline
Glasgow Coma Scale (GCS) score) were assessed at baseline.
OCT, ICH locations (basal ganglia, thalamic, and
lobar), presence of intraventricular hemorrhage (IVH), and
imaging markers were also assessed for analysis. HE was
defined as absolute hematoma growth >6 mL or relative
hematoma growth >33% from initial to follow-up CT
scans.8
Patient selection
The inclusion criteria were as follows: (1) spontaneous
supratentorial ICH; (2) OCT ⩽6 h; and (3) follow-up CT
within 48 h of initial ictus. We excluded patients who met
any of the following criteria: (1) multiple hematomas; (2)
primary IVH; (3) surgical hematoma evacuation before the
follow-up CT; (4) secondary ICH ascribed to trauma, aneurysm,
vascular malformation, moyamoya disease, brain
tumor, hemorrhagic transformation of a cerebral infraction;
(5) prior anticoagulant therapy; (6) abnormal coagulation at
admission (laboratory values: international normalized
ratio >1.7, platelet count <50 × 109/μL); (7) unclear onset;
(8) symptom onset to baseline CT time >6 h; (9) follow-up
CT time >48 h; (10) missing follow-up CT; (11) severe
artifacts on CT imaging during examinations; and (12)
infratentorial (brainstem and cerebellum) ICH.
Neuroimaging
Imaging markers were classified into two categories based
on density and shape features. Considerable overlap was
observed between different types.4 Three markers were used
International Journal of Stroke, 19(2)
WSO - February 2024

Table of Contents for the Digital Edition of WSO - February 2024
