Machine learning integration of routine inflammatory biomarkers for predicting remote punctate ischemic lesions following intracerebral hemorrhage: a single-center retrospective study - Takeaways - MDSpire

Machine learning integration of routine inflammatory biomarkers for predicting remote punctate ischemic lesions following intracerebral hemorrhage: a single-center retrospective study

  • By

  • Yibo Dong

  • Longyun Yi

  • Hongbo Tu

  • June 29, 2026

  • 0 min

Share

  • 1

    Remote Punctate Ischemic Lesions (RPIL) after intracerebral hemorrhage (ICH) are linked to worse functional outcomes.

  • 2

    The study analyzed 12,327 ICH patients, with 6,134 meeting inclusion criteria after applying strict exclusions.

  • 3

    XGBoost was the most effective machine learning model for predicting RPIL, achieving an AUC of 0.799.

  • 4

    Key predictors identified included Age, History of Diabetes, SII, D-Dimer, Glucose, and Fibrinogen.

  • 5

    The study emphasizes the need for multicenter validation to confirm the generalizability of the predictive models.

Original Source(s)

Machine learning integration of routine inflammatory biomarkers for predicting remote punctate ischemic lesions following intracerebral hemorrhage: a single-center retrospective study

  • by Yibo Dong, Longyun Yi, Hongbo Tu

  • June 29, 2026

Related Content

A spatiotemporal state-inference framework for adaptive immunotherapy in glioblastoma

Distress Tied to Digital Overuse

In a survey of 420 Italian adults, psychological distress showed stronger associations than autistic traits with problematic internet and mobile phone use, although both were associated with higher digital-use scores.

From severity scoring to predictive analytics: the emerging role of AI in neurosurgery