Texture analysis for the assessment of structural changes in parotid glands induced by radiotherapy

• Published in: Radiotherapy and oncology Vol. 109; no. 3; pp. 384 - 387
• Main Authors: Scalco, Elisa, Fiorino, Claudio, Cattaneo, Giovanni Mauro, Sanguineti, Giuseppe, Rizzo, Giovanna
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier Ireland Ltd 01.12.2013
• Subjects: Adaptive radiotherapy; Adult; Aged; Carcinoma; Female; Fractal dimension; Hematology, Oncology and Palliative Medicine; Humans; Male; Middle Aged; Nasopharyngeal Carcinoma; Nasopharyngeal Neoplasms - radiotherapy; Parotid Gland - pathology; Parotid Gland - radiation effects; Parotid gland shrinkage; Radiation Injuries - etiology; Radiation Injuries - pathology; Radiotherapy, Intensity-Modulated - adverse effects; Radiotherapy, Intensity-Modulated - methods; Texture analysis; Xerostomia - etiology; Xerostomia - pathology; Parotid gland shrinkage; Texture analysis; Fractal dimension; Adaptive radiotherapy
• ISSN: 0167-8140; 1879-0887
• DOI: 10.1016/j.radonc.2013.09.019

Abstract Background and purpose During radiotherapy (RT) for head-and-neck cancer, parotid glands undergo significant anatomic, functional and structural changes which could characterize pre-clinical signs of an increased risk of xerostomia. Texture analysis is proposed to assess structural changes of parotids induced by RT, and to investigate whether early variations of textural parameters (such as mean intensity and fractal dimension) can predict parotid shrinkage at the end of treatment. Material and methods Textural parameters and volumes of 42 parotids from 21 patients treated with intensity-modulated RT for nasopharyngeal cancer were extracted from CT images. To individuate which parameters changed during RT, a Wilcoxon signed-rank test between textural indices (first and second RT week; first and last RT week) was performed. Discriminant analysis was applied to variations of these parameters in the first two weeks of RT to assess their power in predicting parotid shrinkage at the end of RT. Results A significant decrease in mean intensity (1.7 HU and 3.8 HU after the second and last weeks, respectively) and fractal dimension (0.016 and 0.021) was found. Discriminant analysis, based on volume and fractal dimension, was able to predict the final parotid shrinkage (accuracy of 71.4%). Conclusion Textural features could be used in combination with volume to characterize structural modifications on parotid glands and to predict parotid shrinkage at the end of RT.