BMC Medical Imaging最新文献

Background: The early prediction of malignant cerebral edema (MCE) following endovascular therapy for acute ischemic stroke is of paramount importance for facilitating timely interventions. The present study aimed to create a comprehensive map of lesion topography associated with MCE risk and to build a machine learning model based on these topography-informed radiomics to predict the MCE in stroke patients after endovascular therapy.

Methods: Using voxel-based lesion analyses, we comprehensively quantified the spatial features of infarct location lesions. These topological features were integrated with radiomics to create a hybrid spatial radiomics model. Four machine learning algorithms bases on topography features, radiomics, and Topo-Rad features were developed to predict MCE in acute stroke patients, respectively. The performance of models was evaluated using the receiver operating characteristic curves, decision curve analysis and Net Reclassification Improvement. The SHapley Additive exPlanations (SHAP) method was employed to interpret and visualize the output of the optimal model.

Results: The topography maps for acute stroke patients showed the right temporal lobe and right caudate nucleus were significantly associated with MCE (P < 0.05). For four ML algorithms, the SVM model based on topo-Rad achieved the highest predictive performance (AUC in training/validation set: 0.872/0.842), while no statistically significant difference was observed compared to the model based on topography (0.857/0.812). The SHAP plots demonstrated that the most significant contributors to model performance were related to temporal_pars_of_MCA_R, occipital_pars_of_PCA_R, parietal_pars_of_MCA_R, temporal_pars_of_MCA_L, and parietal_pars_of_MCA_L.

Conclusions: The infarct topography plays a dominant role in predicting MCE following endovascular therapy, with radiomic features providing limited additional predictive value.

Purpose: To construct a radiomics nomogram model predicting the status of lymphovascular tumor embolus (LTE) in patients with lung invasive adenocarcinoma (LAC).

Materials and methods: This retrospective analysis enrolled 195 patients with pathologically-confirmed LAC, treated at Weifang People's Hospital between January 2018 and April 2021, including 152 and 43 cases in the LTE and non-LTE groups, respectively. Regions of interest were manually delineated on preoperative CT images using 3D slicer. Subsequently, 850 radiomics features were extracted and subjected to feature reduction through least absolute shrinkage and selection operator regression. The effectiveness of the predictive model was evaluated using receiver operating characteristic curves, calibration, and decision curve analysis. The log-rank test was applied to data split into low-score and high-score groups to analyze early recurrence-free survival based on the optimal cutoff value established in the mixed model.

Results: Five identified feature parameters were applied to establish a rad-score. Hybrid prediction model integrating smoking status and radiomics signature demonstrated better predictive efficacy than the radiomics models in the training cohort (area under the curve [AUC], 0.9210 vs. 0.8781) and validation cohort (AUC, 0.8807 vs. 0.8770), although without reaching statistical significance. The calibration curves of the nomogram illustrated the goodness-of-fit to predict LTE status in both cohorts. Kaplan-Meier survival curve analysis demonstrated a significant difference in recurrence-free survival rate between the low-score and high-score groups, as predicted based on the optimal cutoff value of the mixed model.

Conclusion: CT radiomics-based model, which could serve as a potential biomarker, demonstrated strong predictive value for LTE status in LAC.