Related Experiment Video
Updated: Jun 28, 2026
07:13
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
AI-assisted Radiomic Model for Cervical Cancer Recurrence Prediction: A Multicenter Retrospective Study with
Shuqing Chen1, Yu Zhang2,3, Dong Chen2
1Department of Radiology, Funan County People's Hospital, Fuyang, Anhui, PR China.
Radiology. Imaging Cancer
|June 26, 2026
Summary
This study developed a clinical-radiomic nomogram using MRI features and biomarkers to predict cervical cancer recurrence, offering accurate, noninvasive prognosis for patients. The model identified key prognostic factors and a biological pathway, improving disease-free survival prediction.
Area of Science:
- Oncology
- Radiology
- Biomedical Engineering
Background:
- Cervical cancer recurrence poses a significant challenge in patient management.
- Accurate prediction of postoperative disease-free survival (DFS) is crucial for personalized treatment strategies.
- Existing prognostic models may not fully integrate diverse data types for comprehensive prediction.
Purpose of the Study:
- To develop and validate a nomogram model for predicting postoperative DFS in cervical cancer patients.
- To incorporate clinical parameters, hematologic inflammatory biomarkers, and MRI radiomic features into the predictive model.
- To explore the underlying biologic mechanism of the identified radiomic signature.
Main Methods:
- A multicenter retrospective study of 804 cervical cancer patients (2016-2023).
- Extraction of 3D radiomic features from pretreatment MRI tumor and peritumoral regions.
- Development of a nomogram using multivariable Cox regression, integrating clinical data, inflammatory markers, and a machine learning-derived radiomics score (Radscore).
- Bioinformatic analysis and in vitro experiments for biologic mechanism validation.
Main Results:
- The integrated clinical-radiomic nomogram demonstrated high predictive performance for 1-, 3-, and 5-year DFS in the external test set (AUCs ranging from 0.84 to 0.93).
- International Federation of Gynecology and Obstetrics stage, squamous cell carcinoma antigen, systemic inflammation response index, and Radscore were independent prognostic factors.
- A TRIM29-cell cycle regulatory axis was identified as the radiomic signature's biologic mechanism.
Conclusions:
- The developed integrated clinical-radiomic nomogram provides accurate, noninvasive prediction of postoperative recurrence in cervical cancer.
- This tool can aid in risk stratification and personalized management of cervical cancer patients.
- The study elucidates a potential biological pathway linked to radiomic features, offering insights for future therapeutic targets.
