Histogram analysis of apparent diffusion coefficient maps for the differentiation between lymphoma and metastatic lymph nodes of squamous cell carcinoma in head and neck region

摘要

Background To clarify the nature of cervical malignant lymphadenopathy is highly important for the diagnosis and differential diagnosis of head and neck tumors. Purpose To investigate the role of first-order apparent diffusion coefficient (ADC) histogram analysis for differentiating lymphoma from metastatic lymph nodes of squamous cell carcinoma (SCC) in the head and neck region. Material and Methods Diffusion-weighted imaging (DWI) data of 67 patients (lymphoma, n?=?20; SCC, n?=?47) with malignant lymphadenopathy were retrospectively analyzed. The SCC group was divided into nasopharyngeal SCC and non-nasopharyngeal SCC groups. The ADC histogram features (ADC_(10), ADC_(25), ADC_(mean), ADC_(median), ADC_(75), ADC_(90), skewness, and kurtosis) were derived and then compared by independent-samples t-test and one-way analysis of variance test, respectively. Receiver operating characteristic curve analyses were employed to investigate diagnostic performance of the significant parameters. Results Lymphoma showed significantly lower ADC_(mean), ADC_(median), ADC_(75), and ADC_(90)than SCC (all P ? ?0.05). Lymphoma showed significantly lower ADC_(25), ADC_(mean), ADC_(median), ADC_(75), and ADC_(90)than non-nasopharyngeal SCC (all P ?