Exploring Applications of Radiomics in Magnetic Resonance Imaging of Head and Neck Cancer: A Systematic Review

Background: Radiomics has been widely investigated for non-invasive acquisition of quantita-tive textural information from anatomic structures. While the vast majority of radiomic analysis is performed on images obtained from computed tomography (CT), magnetic resonance imaging (MRI)-based radiomics has generated increased attention. In head and neck cancer (HNC), how-ever, attempts to perform consistent investigations are sparse, and it is unclear whether the result-ing textural features can be reproduced. To address this unmet need, we systematically reviewed the quality of existing MRI radiomics research in HNC. Methods: Literature search was conducted in accordance with guidelines established by Pre-ferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). Electronic data-bases were examined from January 1990 through November 2017 for common radiomic key-words. Eligible completed studies were then scored using a standardized checklist that we de-veloped from Enhancing the Quality and Transparency of Health Research (EQUATOR) guide-lines for reporting machine learning predictive model specifications and results in biomedical re-search, defined by Luo et al.(1). Descriptive statistics of checklist scores were populated, and a sub-group analysis of methodology items alone was conducted in comparison to overall scores. Results: Sixteen completed studies and four ongoing trials were selected for inclusion. Of the completed studies, the nasopharynx was the most common site of study (27.5%).MRI modalities varied with only four of the completed studies (25%) extracting radiomic features from a single sequence. Study sample sizes ranged between 13-118 patients (median of 20), and final radiomic signatures ranged from 2-279 features. Analyzed end-points included either segmentation or his-topathological classification parameters (44%) or prognostic and predictive biomarkers (56%) . Liu et al.(39) addressed the highest number of our checklist items (Total Score [TS]: 48), and a sub-group analysis of methodology checklist items alone did not demonstrate any difference in scoring trends between studies (Spearman’s ρ = 0.94 [p <.0001]). Conclusions: Although MRI radiomic applications demonstrate predictive potential in analyzing diverse HNC outcomes, methodological variances preclude accurate and collective interpretation of data.