Machine learning combined with resting-state functional MRI to characterize functional brain differences in post-stroke depression - Takeaways - MDSpire

Machine learning combined with resting-state functional MRI to characterize functional brain differences in post-stroke depression

  • By

  • Yuanxin Shao

  • Chao Liang

  • Dan Xu

  • Yang Zhao

  • Phi Thi Thanh Hoa

  • Xue Zhang

  • Dongyang Shi

  • Weifeng Guo

  • June 22, 2026

  • 0 min

Share

  • 1

    Post-stroke depression (PSD) affects 30%-50% of stroke survivors and is linked to poorer rehabilitation outcomes and increased mortality risk.

  • 2

    The study involved 50 PSD patients and 50 healthy controls using resting-state functional MRI to identify brain function variations.

  • 3

    Twenty-nine candidate imaging features were identified, with LASSO regression retaining 10 core features for machine learning analysis.

  • 4

    The Extra Trees model achieved the highest performance with an AUC of 0.889, indicating effective classification of PSD.

  • 5

    SHAP analysis revealed key features associated with PSD, including connectivity and activity in specific brain regions.

Original Source(s)

Machine learning combined with resting-state functional MRI to characterize functional brain differences in post-stroke depression

  • by Yuanxin Shao, Chao Liang, Dan Xu, Yang Zhao, Phi Thi Thanh Hoa, Xue Zhang, Dongyang Shi, Weifeng Guo

  • June 22, 2026

Related Content

Enhancing motor function after stroke: a systematic review and meta-analysis of bioelectrical feedback interventions

A review of the impact of circadian rhythm on motor function in stroke recovery patients

Suicidal ideation after stroke: why standardized screening is essential for community-based prediction models