Robustness of radiomics to variations in segmentation methods in multimodal brain MRI

• Published in: Scientific reports 2022-10, Vol.12 (1), p.16712-16712, Article 16712
• Main Authors: Poirot, M. G., Caan, M. W. A., Ruhe, H. G., Bjørnerud, A., Groote, I., Reneman, L., Marquering, H. A.
• Format: Article
• Language: English
• Subjects: 631/114; 631/378; 631/378/116; 631/67; 692/308/53; 692/308/575; 692/53; 692/53/2423; 692/700; 692/700/1421; 692/700/1421/1628; Classification; Correlation coefficient; Deep learning; Humanities and Social Sciences; Image processing; multidisciplinary; Neuroimaging; Radiomics; Reproducibility; Research methods; Science; Science (multidisciplinary); Segmentation; Sleep deprivation
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/s41598-022-20703-9

Radiomics in neuroimaging uses fully automatic segmentation to delineate the anatomical areas for which radiomic features are computed. However, differences among these segmentation methods affect radiomic features to an unknown extent. A scan-rescan dataset (n = 46) of T1-weighted and diffusion tensor images was used. Subjects were split into a sleep-deprivation and a control group. Scans were segmented using four segmentation methods from which radiomic features were computed. First, we measured segmentation agreement using the Dice-coefficient. Second, robustness and reproducibility of radiomic features were measured using the intraclass correlation coefficient (ICC). Last, difference in predictive power was assessed using the Friedman-test on performance in a radiomics-based sleep deprivation classification application. Segmentation agreement was generally high (interquartile range = 0.77–0.90) and median feature robustness to segmentation method variation was higher (ICC > 0.7) than scan-rescan reproducibility (ICC 0.3–0.8). However, classification performance differed significantly among segmentation methods (p