Establishing ADC-Based Histogram and Texture Features for Early Treatment-Induced Changes in Head and Neck Squamous Cell Carcinoma.

Anna Rodrigues,Taofik Oyekunle,Jeff Nawrocki,Donna Niedzwiecki,Zheng Chang,Yvonne M Mowery,David M Brizel,Jenny K Hoang,Oana Craciunescu,Kelly Loman

doi:10.3389/fonc.2021.708398

Anna Rodrigues, Taofik Oyekunle + Show 8 more

Open Access

https://doi.org/10.3389/fonc.2021.708398

Copy DOI

Abstract

The purpose of this study was to assess baseline variability in histogram and texture features derived from apparent diffusion coefficient (ADC) maps from diffusion-weighted MRI (DW-MRI) examinations and to identify early treatment-induced changes to these features in patients with head and neck squamous cell carcinoma (HNSCC) undergoing definitive chemoradiation. Patients with American Joint Committee on Cancer Stage III–IV (7th edition) HNSCC were prospectively enrolled on an IRB-approved study to undergo two pre-treatment baseline DW-MRI examinations, performed 1 week apart, and a third early intra-treatment DW-MRI examination during the second week of chemoradiation. Forty texture and six histogram features were derived from ADC maps. Repeatability of the features from the baseline ADC maps was assessed with the intra-class correlation coefficient (ICC). A Wilcoxon signed-rank test compared average baseline and early treatment feature changes. Data from nine patients were used for this study. Comparison of the two baseline ADC maps yielded 11 features with an ICC ≥ 0.80, indicating that these features had excellent repeatability: Run Gray-Level Non-Uniformity, Coarseness, Long Zone High Gray-Level, Variance (Histogram Feature), Cluster Shade, Long Zone, Variance (Texture Feature), Run Length Non-Uniformity, Correlation, Cluster Tendency, and ADC Median. The Wilcoxon signed-rank test resulted in four features with significantly different early treatment-induced changes compared to the baseline values: Run Gray-Level Non-Uniformity (p = 0.005), Run Length Non-Uniformity (p = 0.005), Coarseness (p = 0.006), and Variance (Histogram) (p = 0.006). The feasibility of histogram and texture analysis as a potential biomarker is dependent on the baseline variability of each metric, which disqualifies many features.

Highlights

Recent publications have summarized the current state of radiomics research utilizing functional imaging in head and neck cancer for tumor segmentation, prognostic and predictive response biomarkers, and monitoring of normal tissue sequelae [1, 2]
Diffusion-weighted magnetic resonance imaging (DW-MRI)-derived apparent diffusion coefficient (ADC) maps express water molecular motion, which tends to be relatively low in tumors due to their higher cellular density than normal tissue
In a similar study by Kim et al that included early intra-treatment examinations (1 week after treatment began), the change in ADC between baseline and intra-treatment examinations had the highest sensitivity for differentiating complete and partial responders [11]. These studies have shown the viability of using DW-MRI in the setting of detecting response during or shortly after completion of treatment by utilizing first-order histogram imaging features

Summary

Introduction

Recent publications have summarized the current state of radiomics research utilizing functional imaging in head and neck cancer for tumor segmentation, prognostic and predictive response biomarkers, and monitoring of normal tissue sequelae [1, 2]. Vandecaveye et al determined that percentage change in ADC from 3 weeks post-CRT relative to baseline examination was significantly correlated with patient outcome, allowing for early assessment of treatment response [12]. In a similar study by Kim et al that included early intra-treatment examinations (1 week after treatment began), the change in ADC between baseline and intra-treatment examinations had the highest sensitivity for differentiating complete and partial responders [11]. These studies have shown the viability of using DW-MRI in the setting of detecting response during or shortly after completion of treatment by utilizing first-order histogram imaging features. Beyond the use of first-order ADC histogram features such as mean ADC, recent studies have evaluated the use of secondorder texture features for response prediction [1, 2, 14, 15]

Objectives

Methods

Results