peritumoral regions

This study aimed to evaluate the stability and reproducibility of CT radiomic features extracted from the peritumoral regions of lung lesions. Stability was defined as the consistency of a feature across different segmentations, while reproducibility was defined as the consistency of a feature across different image acquisitions. The study utilized two datasets to measure stability and reproducibility, and a total of 264 statistical, histogram, and texture radiomic features were assessed from the selected peritumoral ROIs.

The results showed that statistical, histogram and a subset of texture features were stable and reproducible, regardless of the peritumoral distances, suggesting that these features are less affected by changes in the peritumoral region due to different segmentations and image acquisitions. However, the stability and reproducibility of Laws and wavelet texture features were inconsistent across all peritumoral distances. The study also found that a subset of features was consistently stable irrespective of the initial parameters for a given segmentation algorithm, and no significant differences were found in stability for features extracted from ROIs bounded by a lung parenchyma mask versus those that were not.

Furthermore, the clinical utility of stable and reproducible peritumoral features was tested in three previously published lung cancer datasets using overall survival as the endpoint. The results showed that stable and reproducible features were more likely to create repeatable models than unstable and nonreproducible features.