Photo Credit: DouglasOlivares

The following is a summary of “Multimodal MRI radiomics-based stacking ensemble learning model with automatic segmentation for prognostic prediction of HIFU ablation of uterine fibroids: a multicenter study,” published in the December 2024 issue of Obstetrics & Gynecology by Wen et al. 

Researchers conducted a retrospective study to assess an MRI radiomics stacking ensemble model combining T2-weighted imaging (T2WI) and contrast-enhanced T1-weighted imaging (CE-T1WI) with deep learning (DL)-based segmentation for predicting high-intensity focused ultrasound (HIFU) ablation outcomes in uterine fibroids. 

They collected data from 360 patients with uterine fibroids who underwent HIFU treatment, with 240 in the training set, 60 in the internal test set, and 60 in the external test set. Automated segmentation was performed using a V-net DL model. Radiomics features from T2WI and CE-T1WI were extracted and preprocessed. Feature selection was done using t-test, Pearson correlation, and LASSO. Support Vector Machine (SVM), Random Forest (RF), Light Gradient Boosting Machine (LightGBM), and Multilayer Perceptron (MLP) were used as base learners to construct predictive models, which were integrated into a stacking ensemble model with Logistic Regression as the meta-learner. Model performance was assessed using an area under the curve (AUC). 

The results showed that among the base models, the MLP achieved an AUC of 0.858 (95% CI: 0.756–0.959) in the internal test set and 0.828 (95% CI: 0.726–0.930) in the external test set. The SVM, LightGBM, and RF models had AUC values of 0.841 (95% CI: 0.737–0.946), 0.823 (95% CI: 0.711–0.934), and 0.750 (95% CI: 0.619–0.881), respectively. The stacking ensemble model showed an AUC of 0.897 (95% CI: 0.818–0.977) in the internal test set and 0.854 (95% CI: 0.759–0.948) in the external test set. 

Investigators found that the DL-based MRI radiomics stacking ensemble model accurately predicted HIFU ablation outcomes for uterine fibroids, outperforming individual models. 

Source: frontiersin.org/journals/physiology/articles/10.3389/fphys.2024.1507986/full