Benchmarking Eliminative Radiomic Feature Selection for Head and Neck Lymph Node Classification

Abstract

In head and neck squamous cell carcinoma (HNSCC) pathologic cervical lymph nodes (LN) remain important negative predictors. Current criteria for LN-classification in contrast-enhanced computed-tomography scans (contrast-CT) are shape-based; contrast-CT imagery allows extraction of additional quantitative data (“features”). The data-driven technique to extract, process, and analyze features from contrast-CTs is termed “radiomics”. Extracted features from contrast-CTs at various levels are typically redundant and correlated. Current sets of features for LN-classification are too complex for clinical application. Effective eliminative feature selection (EFS) is a crucial preprocessing step to reduce the complexity of sets identified. We aimed at exploring EFS-algorithms for their potential to identify sets of features, which were as small as feasible and yet retained as much accuracy as possible for LN-classification. In this retrospective cohort-study, which adhered to the STROBE guidelines, in total 252 LNs were classified as “non-pathologic” (n = 70), “pathologic” (n = 182) or “pathologic with extracapsular spread” (n = 52) by two experienced head-and-neck radiologists based on established criteria which served as a reference. The combination of sparse discriminant analysis and genetic optimization retained up to 90% of the classification accuracy with only 10% of the original numbers of features. From a clinical perspective, the selected features appeared plausible and potentially capable of correctly classifying LNs. Both the identified EFS-algorithm and the identified features need further exploration to assess their potential to prospectively classify LNs in HNSCC.

Current Applications of Artificial Intelligence to Classify Cervical Lymph Nodes in Patients with Head and Neck Squamous Cell Carcinoma—A Systematic Review

“…Locally-advanced HNSCC patients require multimodality treatment [ 2 ], frequently consisting of surgery followed by radiotherapy or primary concurrent radiochemotherapy [ 3 , 4 ]. The presence and the localization of pathologic cervical LNs with or without extracapsular spread (ECS) remain important negative prognostic factors in such patients in terms of overall survival [ 5 , 6 , 7 , 8 ].…”

“…Mainly, and without further quantification, head and neck radiologists consider LN maximum diameter and margins. Consequently, prognostic implications previously observed for pathologic cervical LNs [ 5 , 6 , 7 , 8 ] are based on these shaped-based detection and classifications criteria [ 14 , 15 , 16 , 17 , 18 , 19 ]. However, significantly more quantitative information about shape, texture, and intensity is contained within CTs, MRIs, and [ 18 F]FDG-PET that could be exploited for detection, segmentation, classification, and exploration of prognostic implications of LNs in patients with locally-advanced HNSCC [ 5 ], thus, ultimately aiding in clinical decision-making [ 5 ].…”

“…Radiomic analysis [ 20 , 21 , 22 ] is an emerging, data-driven technique aiming at extracting, processing, and analyzing such quantitative information from images. The core idea is to treat imaging data, which is frequently obtained during routine clinical staging, as data-mineable sources of information to aid in these processes [ 5 ]. This quantitative information contained within images represents a wealth of data for analysis using artificial intelligence (AI).…”

“…The typical radiomic analysis workflow in the context of cervical LNs in locally-advanced HNSCC patients starts with detection, followed by segmentation and classification based on the extracted shape, intensity, and texture features, before prognostic implications are explored [ 5 ]. Thus, AI and its respective subsets ML, ANN, and DL may be applied to all of these tasks, ultimately aiming at improving the efficacy of the workflow [ 5 , 26 ].…”

MRI-based radiomics analysis for preoperative evaluation of lymph node metastasis in hypopharyngeal squamous cell carcinoma

“…Recently, Radiomics features extracted from primary tumors exhibited great potential in prediction of cervical lymph node metastasis (13,14). However, only a few studies have explored the role of radiomics features in lymph node evaluation of head and neck cancers (15,16). To our knowledge, there is no radiomics study for prediction of cervical lymph node metastasis in HPSCC.…”