EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Acute Ischemic Stroke Lesion Characterization: Comparative Analysis of Cortical and Basal Ganglia Lesions Using Time-Histogram Metrics

Hon-Man Liu iD, Wei-Lung Tseng iD

Diagnostics · 06.05.2026 · 16 (9) · 1400

DOI: 10.3390/diagnostics16091400 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Diagnostics, online erschienen am 06.05.2026. Verfasst von Hon-Man Liu und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Medical Imaging, Fu Jen Catholic University Hospital, 69, Guizi Road, Taishan District, New Taipei City 24352, Taiwan. Die Arbeit nennt 16 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Fu-Jen Catholic University Hospital., Fu Jen Catholic University Hospital.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background and Purpose: Understanding regional differences in acute ischemic stroke (AIS) lesions, particularly between cortical and basal ganglia (BG) regions, is crucial for enhancing diagnostic precision and therapeutic strategies. This study builds upon prior research in CT-based stroke imaging by introducing entropy and SD as novel biomarkers for lesion differentiation and focuses on their temporal evolution to identify critical diagnostic windows for clinical management.

Methods

We retrospectively analyzed imaging data from 45 AIS patients (27 cortical, 18 BG lesions). Time-histogram metrics—standard deviation (SD), skewness, kurtosis, and entropy—were computed for lesions in both regions across multiple time points post-stroke. Statistical comparisons utilized one-way ANOVA and post hoc t-tests (p < 0.05).

Results

Cortical lesions exhibited significantly higher entropy (4.34 vs. 3.99, p = 0.000034) and SD (5.16 vs. 3.94, p = 0.000144) compared to BG lesions, reflecting greater heterogeneity. No significant differences were found in skewness or kurtosis. Peak diagnostic sensitivity occurred at 76–87 min post-stroke (p < 0.001). Temporal trends revealed increasing divergence in entropy and SD between cortical and BG lesions during this window. The identification of the 76–87-min diagnostic window offers critical insights into the hyperacute phase, where early ischemic changes are often subtle on NCCT.

Conclusions

SD and entropy are robust biomarkers for distinguishing cortical and BG lesions in AIS, offering insights into regional tissue responses and temporal evolution, with potential for personalized stroke care. These metrics could serve as standalone biomarkers or be integrated into AI-based NCCT triage systems to enhance lesion characterization and therapeutic decision-making.

Schlagwörter

Stichwörter der Autor:innen

acute ischemic strokebasal ganglia lesioncortical lesionstatistical metricstime histogram

Autor:innen und Institutionen

  • Department of Medical Imaging, Fu Jen Catholic University Hospital, 69, Guizi Road, Taishan District, New Taipei City 24352, Taiwan.
  • Department of Medical Imaging, National Taiwan University Hospital, Taipei 110, Taiwan.
  • Department of Neurosurgery, Fu Jen Catholic University Hospital, New Taipei City 24352, Taiwan.

ORCID-Profile: Hon-Man Liu, Wei-Lung Tseng

Förderung

  • Fu-Jen Catholic University Hospital.
  • Fu Jen Catholic University Hospital

Zitieren

APA

Liu, H. M., & Tseng, W. L. (2026). Acute Ischemic Stroke Lesion Characterization: Comparative Analysis of Cortical and Basal Ganglia Lesions Using Time-Histogram Metrics. Diagnostics, 16(9), 1400. https://doi.org/10.3390/diagnostics16091400

Vancouver

Liu HM, Tseng WL. Acute Ischemic Stroke Lesion Characterization: Comparative Analysis of Cortical and Basal Ganglia Lesions Using Time-Histogram Metrics. Diagnostics. 2026;16(9):1400. doi:10.3390/diagnostics16091400

BibTeX
@article{Liu2026,
  author = {Liu, Hon-Man and Tseng, Wei-Lung},
  title = {Acute Ischemic Stroke Lesion Characterization: Comparative Analysis of Cortical and Basal Ganglia Lesions Using Time-Histogram Metrics},
  journal = {Diagnostics},
  year = {2026},
  volume = {16},
  number = {9},
  pages = {1400},
  doi = {10.3390/diagnostics16091400}
}
Kostenlos starten →