Related Experiment Video
Updated: Jun 4, 2025
06:24
Dynamic Contrast Enhanced Magnetic Resonance Imaging of an Orthotopic Pancreatic Cancer Mouse Model
Published on: April 18, 2015
15.1K
Delta-Radiomics Approach Using Contrast-Enhanced and Noncontrast-Enhanced Computed Tomography Images for Predicting
Takanori Adachi1, Mitsuhiro Nakamura1,2, Takahiro Iwai1
1Department of Radiation Oncology and Image-Applied Therapy, Kyoto University, Shogoin, Sakyo-ku, Kyoto, Japan.
Advances in Radiation Oncology
|December 17, 2024
Summary
Predicting distant metastasis in borderline resectable pancreatic cancer is possible using delta-radiomics from CT scans. This method, utilizing differences between contrast-enhanced and non-contrast CT images, aids in risk stratification for patients.
Area of Science:
- Oncology
- Radiology
- Data Science
Background:
- Pancreatic cancer poses a significant challenge due to its high mortality rate.
- Predicting distant metastasis (DM) is crucial for treatment planning in borderline resectable pancreatic carcinoma.
- Current prediction methods may not fully leverage imaging data.
Purpose of the Study:
- To predict distant metastasis (DM) in patients with borderline resectable pancreatic carcinoma.
- To evaluate the efficacy of delta-radiomics features derived from contrast-enhanced computed tomography (CECT) and non-CECT images.
- To compare the predictive performance of clinical, radiomics, and hybrid models.
Main Methods:
- Utilized data from 67 eligible patients with borderline resectable pancreatic carcinoma.
- Extracted 3906 radiomics features from CECT and non-CECT images, calculating delta-radiomics features.
- Developed predictive models (clinical, radiomics, hybrid) using Fine-Gray regression and random survival forest, validated on a test set.
Main Results:
- The random survival forest (RSF) model incorporating delta-radiomics achieved the highest predictive performance (concordance index: 0.727).
- The RSF-based model with delta-radiomics significantly stratified patients into high- and low-risk groups for DM (P < .05).
- Specific delta-radiomics features, particularly from the gray-level size-zone matrix, correlated with DM incidence.
Conclusions:
- Delta-radiomics features derived from CECT and non-CECT images, when analyzed with RSF, are effective in predicting DM in borderline resectable pancreatic carcinoma.
- This approach offers a promising tool for improved risk assessment and personalized treatment strategies.
