An area under the curve (AUC) ≥ 0.60 was chosen to retain potential features. The first step selected features based on their predictive value of the pCR status.
The feature set reduction workflow was developed as a two-step approach. Clinical and radiomic features (either original or harmonized) were then selected in the training set only. Patients from institution 1 were included in the training cohort while patients from institution 2 formed the testing cohort.
Conclusions: After correction of inter-site variability and imbalanced data, addition of radiomic features enhances the prediction of pCR after neoadjuvant CRT in LARC. After ComBat harmonization, the radiomic and the combined models obtained a Bacc of 68.2% and 85.5%, respectively, while the clinical model and the pre-ComBat combined achieved respective Baccs of 60.0% and 75.5%. Results: Out of the 124 included patients, 14 had pCR (11.3%). Each model was then evaluated on the testing set using sensitivity, specificity, balanced accuracy (Bacc) with the predefined cut-off. A cut-off maximizing the model’s performance was defined on the training set. Five pCR prediction models (clinical, radiomics before and after ComBat, and combined before and after ComBat) were then developed on the training set with a neural network approach and a bootstrap internal validation ( n = 1000 replications). We selected the most predictive characteristics using Spearman’s rank correlation and the Area Under the ROC Curve (AUC). The ComBat and Synthetic Minority Over-sampling Technique (SMOTE) approaches were used to account for inter-institution heterogeneity and imbalanced data, respectively. The cohort was split into training (Institution 1) and testing (Institution 2) sets. In total, 88 radiomic parameters were extracted from each sequence using the Miras© software, with a total of 822 features by patient. The tumor was manually segmented on the T2-weighted and diffusion axial MRI sequences and on CE-CT. Both pre-CRT pelvic MRI and CE-CT were mandatory for inclusion. Material: All patients treated for a LARC with neoadjuvant CRT and subsequent surgery in two separate institutions between 20 were considered. Objective: Our objective was to develop a radiomics model based on magnetic resonance imaging (MRI) and contrast-enhanced computed tomography (CE-CT) to predict pathological complete response (pCR) to neoadjuvant treatment in locally advanced rectal cancer (LARC).
0 kommentar(er)
