Magnetic Resonance Imaging Texture Analysis in the Detection of Metastatic Lymph Nodes in Patients with Nasopharyngeal Carcinoma

Authors : Halil ÖZER, Abdussamet BATUR, Nurullah ÖZDEMİR, Mehmet Sedat DURMAZ, Abidin KILINÇER

Pages : 461-465

Doi:10.54005/geneltip.1311577

View : 57 | Download : 64

Publication Date : 2023-08-31

Article Type : Research Paper

Abstract :Abstract Aims: To investigate the role of magnetic resonance imaging insert ignore into journalissuearticles values(MRI); texture analysis insert ignore into journalissuearticles values(TA); in the detection of metastatic lymph nodes in patients with nasopharyngeal carcinoma insert ignore into journalissuearticles values(NPC);. Material and methods: Between January 2020 and October 2021, 15 NPC patients with 32 metastatic lymph nodes and 30 healthy subjects with benign lymph nodes were included in the study. The texture features compared between metastatic and benign lymph nodes. The independent predictor parameters of metastatic lymph nodes were determined using multivariate regression analysis. Receiver operator characteristics insert ignore into journalissuearticles values(ROC); analysis was used to evaluate the diagnostic performance of the regression models. Results: The first order texture features did not differ significantly between groups insert ignore into journalissuearticles values(p>0.05);. Except for correlation in metastatic lymph nodes, all gray-level co-occurrence matrix insert ignore into journalissuearticles values(GLCM); and gray-level run length matrix insert ignore into journalissuearticles values(GLRLM); features were significantly different insert ignore into journalissuearticles values(p<0.05);. The GLCM features of joint entropy, joint energy, and maximum probability; and the GLRLM features of gray level non uniformity and low gray level run emphasis were independent predictors of metastatic lymph nodes. The area under the curve insert ignore into journalissuearticles values(AUC); values for the GLCM regression model and GLRLM regression model were 0.975 and 0.928, respectively. Conclusion: MRI texture analysis may be useful to detect metastatic lymph nodes in patients with NPC by providing quantitative information on tissue heterogeneity and cellular composition.
Keywords : Manyetik rezonans görüntüleme, doku analizi, nazofarenks karsinomu, metastatik lenf nodları