Quantitative CT Imaging Radiomics-Based Prediction of Bone Mineral Density Changes After Sleeve Gastrectomy
Deyao Hu1, Jilu Ruan1, Chengjian Liu1
1Department of Medical Imaging, Qujing Second People's Hospital, Qujing City, Yunnan Province, Qujing, China.
Obesity Surgery
|June 19, 2025
Summary
Preoperative vertebral quantitative CT (QCT) radiomic features effectively predict bone mineral density (BMD) changes after sleeve gastrectomy (SG). Combining QCT with clinical data offers the best prediction accuracy for postoperative BMD.
Area of Science:
- Radiology
- Oncology
- Medical Imaging
Background:
- Sleeve gastrectomy (SG) can impact bone mineral density (BMD).
- Predicting postoperative BMD changes is crucial for patient management.
- Quantitative CT (QCT) radiomics offers potential for non-invasive assessment.
Purpose of the Study:
- To extract preoperative vertebral QCT radiomic features.
- To develop predictive models for BMD changes post-SG.
- To evaluate the performance of clinical, radiomic, and combined models.
Main Methods:
- Retrospective analysis of 203 SG patients.
- Feature extraction, normalization, and selection (ICC, LASSO).
- XGBoost model development and evaluation (ROC, AUC, DCA).
Main Results:
- Significant differences in BMI, erector spinae CT value, AST, and ALT observed.
- High AUC values achieved by radiomic (0.96) and combined (0.96) models in the test set.
- Combined model demonstrated the highest net benefit via DCA.
Conclusions:
- Vertebral QCT radiomics is a valuable tool for predicting BMD changes after SG.
- Integration of QCT radiomic and clinical features enhances predictive accuracy.
- This approach aids in personalized patient care following bariatric surgery.
