Histogram analysis of synthetic magnetic resonance imaging: Correlations with histopathological factors in head and neck squamous cell carcinoma. Issue 160 (March 2023)

• Record Type: Journal Article
• Authors: Yang, Fan; Li, Yujie; Lei, Huizi; Wei, Haoran; Du, Qiang; Yu, Xiaoduo; Li, Xiaolu; Zhao, Yanfeng; Li, Lin; Xie, Lizhi; Lin, Meng
• Is Part Of: European journal of radiology. Issue 160(2023)
• Publication Date: 2023-03
• Subjects: Head and neck -- Synthetic MRI -- Histogram analysis -- Histological grade -- Human papillomavirus; AUC Area under the curve -- EGFR Epidermal growth factor receptor -- HNSCC Head and neck squamous cell carcinoma -- LOOCV Leave-one-out cross-validation -- MAD Mean absolute deviation -- PD Proton density -- ROC Receiver operating characteristic -- SyMRI Synthetic magnetic resonance imaging -- T2WI T2-weighted imaging -- VOI Volume of interest; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2023.110715
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: SyMRI histogram parameters could assess different histopathological factors in HNSCC. T2_Entropy showed the highest diagnostic performance in discriminating HPV status. PD map showed good performance in discriminating histological Grade 1/2 from Grade 3. Several SyMRI histogram parameters correlated with p16, Ki-67, EGFR and p53 status. Abstract: Purpose: To analyse the association between histogram parameters derived from synthetic MRI (SyMRI) and different histopathological factors in head and neck squamous cell carcinoma (HNSCC). Method: Sixty-one patients with histologically proven primary HNSCC were prospectively enrolled. The correlations between histogram parameters of SyMRI (T1, T2 and proton density (PD) maps) and histopathological factors were analysed using Spearman analysis. The Mann-Whitney U test or Student's t test was utilized to differentiate histological grades and human papillomavirus (HPV) status. The ROC curves and leave-one-out cross-validation (LOOCV) were used to evaluate the differentiation performance. Bootstrapping was applied to avoid overfitting. Results: Several histogram parameters were associated with histological grade: T1 map (r = 0.291) and PD map (r = 0.294 – 0.382/−0.343), and PD_75th Percentile showed the highest differentiation performance (AUC: 0.721 (ROC) and 0.719 (LOOCV)). Moderately negative correlations were found between p16 status and the histogram parameters: T1 map (r = −0.587 – −0.390), T2 map (r = −0.649 – −0.357) and PD map (r = −0.537 – −0.338). In differentiating HPV infection, Entropy was the most discriminative parameter in each map and T2_Entropy showed the highest diagnostic performance (AUC: 0.851 [ROC] and 0.851 [LOOCV]). Additionally, several histogram parameters were correlated with Ki-67 (r = −0.379/−0.397), epidermal growth factor receptor (EGFR) (r = 0.318/0.322) status and p53 (r = 0.452 – 0.665/−0.607) status. Conclusions: Histogram parameters derived from SyMRI may serve as a potential biomarker for discriminating relevant histopathological features, including histological differentiation grade, HPV infection, Ki-67, EGFR and p53 statuses.