Machine learning-based prediction of survival prognosis in cervical cancer
Problem
Clinical features alone are insufficient to forecast cervical cancer prognosis, so better predictive models are needed to improve patient management.
Why it matters
Accurate prognosis forecasting could improve cervical cancer treatment decisions and patient survival.
What the AI did
A machine learning model used TCGA miRNA expression data; Cox Proportional-Hazards analysis identified 42 survival-related miRNAs and K-means clustering grouped patients by 5-year survival outcome.
Results
42 survival-related miRNAs identified; patients clustered into four groups with 5-year survival outcomes of ≥90%, ≈65%, ≤40%
Impact
Model demonstrated on public TCGA data; clinical adoption not stated in source.
Record
| Venue | BMC Bioinformatics |
|---|---|
| Published | 2021 |
| Work type | article |
| Authors | Dongyan Ding, Tingyuan Lang, Dongling Zou, Jiawei Tan, Jia Chen, Lei Zhou, Dong Wang, Rong Li et al. |
| Citations | 53 (OpenAlex, at import time) |
| References | 57 |
| Model or method | K-means clustering with Cox Proportional-Hazards analysis |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1186/s12859-021-04261-x |
| Open access | gold |
| Keywords | cervical cancer, miRNA, machine learning, survival prediction, K-means clustering |
Importance rating
significant — thousands to millions affected
Better cervical cancer prognosis could improve treatment decisions and survival, a clear significant benefit, but the abstract states no incidence, mortality, or population scale; demonstrated only on TCGA data.