Preoperative MRI radiomic analysis for predicting local tumor progression in colorectal liver metastases before microwave ablation
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.MRI radiomics can predict local tumor progression in colorectal liver metastases (CLM) after microwave ablation (MWA). Combined clinical and radiomic models showed the highest accuracy for predicting progression-free survival.
Area Of Science
- Radiology
- Oncology
- Medical Imaging
Background
- Radiomics shows promise in predicting prognosis for colorectal liver metastases (CLM).
- Existing radiomics data primarily focuses on CT, with limited information available for MRI.
- Accurate prediction of local tumor progression is crucial for effective treatment planning in CLM patients.
Purpose Of The Study
- To assess the capability of MRI-derived radiomic features (RFs) to predict local tumor progression-free survival (LTPFS).
- To evaluate the performance of radiomic models compared to clinical models for LTPFS prediction in CLM patients treated with microwave ablation (MWA).
Main Methods
- Retrospective evaluation of CLM patients treated with MWA and pre-operative Gadoxetic acid-MRI.
- Manual segmentation of CLMs on T2 and T1-Hepatobiliary phase (T1-HBP) MRI scans.
- Development and comparison of clinical (CLIN-mod), radiomic (RAD-T1, RAD-T2), and combined (COMB-T1, COMB-T2) models for LTPFS prediction using Cox multivariate analysis.
Main Results
- Combined models demonstrated superior performance in predicting LTPFS compared to clinical or radiomic models alone.
- COMB-T1 achieved an AUC of 0.98 (p < 0.0001), and COMB-T2 achieved an AUC of 0.95 (p < 0.0003).
- Key predictors in combined models included minimal ablation margins, tumor regression grade, signal intensity percentiles, and tumor flatness.
Conclusions
- MRI-based radiomic evaluation is feasible and valuable for predicting local tumor progression in CLMs.
- Combined clinical and radiomic models offer enhanced predictive power for LTPFS in CLM patients undergoing MWA.
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.