A computed tomography‑based radio‑clinical model for the prediction of microvascular invasion in gastric cancer

Yahan Tong ^1,2, Can Hu ^2,3, Xiaoping Cen ⁴

⁰Department of Radiology, Zhejiang Cancer Hospital, Hangzhou, Zhejiang 310022, P.R. China.

Molecular and Clinical Oncology +

|

November 1, 2024

View abstract on PubMed

Summary

A new radio-clinical model effectively predicts microvascular invasion (MI) in gastric cancer before surgery. Combining radiomics and clinical data, this model improves prediction accuracy, aiding personalized treatment strategies for gastric cancer patients.

Area Of Science

Oncology
Radiology
Medical Imaging

Background

Microvascular invasion (MI) is a critical prognostic factor in gastric cancer.
Accurate pre-operative prediction of MI is essential for treatment planning.

Purpose Of The Study

To develop and validate a radio-clinical model for predicting pre-operative microvascular invasion (MI) in gastric cancer.
To integrate radiological features from CT scans with clinical characteristics.

Main Methods

Retrospective analysis of 534 gastric cancer patients.
Extraction of radiomics features from contrast-enhanced CT images.
Development of a radio-clinical model using logistic regression and validated with ROC and DCA.

Main Results

The radiomics signature model showed moderate predictive ability (AUC 0.73-0.77).
The combined radio-clinical model demonstrated superior performance (AUC 0.80-0.88) in both training and test sets.
Model validation confirmed good fit and clinical applicability.

Conclusions

A pre-operative radio-clinical model integrating radiomics and clinical data can accurately predict MI in gastric cancer.
This model supports personalized treatment strategies for gastric cancer.
The model shows significant potential for clinical application in gastric cancer management.

Keywords:

X-ray computed gastric cancer microvascular invasion nomogram radiomics tomography