Non-invasive prediction of Ki-67 and p53 biomarkers in spinal ependymoma via deep learning: using multimodal magnetic resonance imaging and clinical data
This study developed a deep learning framework for non-invasive prediction of Ki-67 and p53 in spinal ependymomas, integrating multimodal MRI and clinical data. The SegFormer model achieved high-precision segmentation, ensuring robust feature extraction. LGBMNet, combining Multilayer Perceptron and Light Gradient Boosting Machine, demonstrated strong predictive performance. Our results confirm that deep learning can effectively predict tumor biomarkers preoperatively, aiding precision neurosurgery.
Authors
- Fuyou Guo (ORCID: https://orcid.org/0000-0001-9332-8272)
- Yu-Zhe Ying
- Zhe Meng (ORCID: https://orcid.org/0000-0002-9883-6951)
- Lu Yang (ORCID: https://orcid.org/0000-0001-9989-6836)
- Chao Ma (ORCID: https://orcid.org/0000-0002-8339-0176)
- Dengpan Song (ORCID: https://orcid.org/0000-0001-5240-7601)
- Liyang Wang (ORCID: https://orcid.org/0000-0002-6175-9219)
- Kaiyuan Yang (ORCID: https://orcid.org/0000-0002-6593-7836)
- Linkai Jing (ORCID: https://orcid.org/0000-0002-1655-7538)
- Guihuai Wang
Institutions
- Capital Medical University (CN)
- Beijing Chao-Yang Hospital, Capital Medical University (CN)
- Beijing Tsinghua Chang Gung Hospital (CN)
- First Affiliated Hospital of Zhengzhou University (CN)
- Beijing Chaoyang Emergency Medical Center (CN)
- Tsinghua University (CN)
Publication Details
- Journal
- Biomarker Research
- Published
- 2025-12-02
- DOI
- https://doi.org/10.1186/s40364-025-00879-8
- Primary Topic
- Glioma Diagnosis and Treatment
- Type
- article
- Field-Weighted Citation Impact
- 0.00