Learning from scanners: Bias reduction and feature correction in radiomics. (November 2019)

• Record Type: Journal Article
• Authors: Zhovannik, Ivan; Bussink, Johan; Traverso, Alberto; Shi, Zhenwei; Kalendralis, Petros; Wee, Leonard; Dekker, Andre; Fijten, Rianne; Monshouwer, René
• Is Part Of: Clinical and translational radiation oncology. Volume 19(2019)
• Page Start: 33
• Page End: 38
• Publication Date: 2019-11
• Subjects: Radiomics -- Reproducibility -- Predictive modeling -- Phantoms; Cancer -- Radiotherapy -- Periodicals; Oncology -- Periodicals; Cancer -- Radiotherapy; Oncology; Radiation Oncology; Neoplasms -- radiotherapy; Translational Medical Research; Periodicals; Electronic journals; Periodicals; 616.9940642
• DOI: 10.1016/j.ctro.2019.07.003
• Languages: English
• ISSNs: 2405-6308

Abstract

Highlights: Typically encountered CT exposure variations result in radiomics values distortion. Proposed correction method suppresses the non-tumor related variation in radiomics. The clinical relevance was shown using a 221 lung cancer patient cohort. Abstract: Purpose: Radiomics are quantitative features extracted from medical images. Many radiomic features depend not only on tumor properties, but also on non-tumor related factors such as scanner signal-to-noise ratio (SNR), reconstruction kernel and other image acquisition settings. This causes undesirable value variations in the features and reduces the performance of prediction models. In this paper, we investigate whether we can use phantom measurements to characterize and correct for the scanner SNR dependence. Methods: We used a phantom with 17 regions of interest (ROI) to investigate the influence of different SNR values. CT scans were acquired with 9 different exposure settings. We developed an additive correction model to reduce scanner SNR influence. Results: Sixty-two of 92 radiomic features showed high variance due to the scanner SNR. Of these 62 features, 47 showed at least a factor 2 significant standard deviation reduction by using the additive correction model. We assessed the clinical relevance of radiomics instability by using a 221 NSCLC patient cohort measured with the same scanner. Conclusions: Phantom measurements show that roughly two third of the radiomic features depend on the exposure setting of the scanner. The dependence can be modeled and corrected significantly reducing the variation in feature values with at least a factor of 2. More complex models will likely increase the correctability. Scanner SNR correction will result in more reliable radiomics predictions in NSCLC.