Title
The Tract-based Spatial Statistics analysis for radiation-induced brain injury in head and neck tumor treatment
Introduction
In head neck tumor treatment, external beam radiotherapy (EBRT) is a conventional technique to destroy cancer cells while at the same time it frequently damage healthy cells and tissues. We aim to study the side effect by exploring accurate measurements along white matter tracts of normal tissue area during radiation treatment by diffusion tensor imaging technique.Questions
What side effect does the radiotherapy bring on the white matter tracts of normal tissue area?Methods
We have 202 patients with head and neck tumors. For each patient we took his T1, Diffusion weighted image (DWI) and CT image before radiotherapy, T1 and DWI image after his radiotherapy, and a 3D dose image during the radiotherapy. The gross tumor volume (GCV) for each CT image was segmented out by our oncologists. The demographic variables for all individuals include, age, gender, race, radiotherapy start and end date, the Karnofsky Performance Status Scale Index, patients' survival days, dose fraction and chemotherapy type. We develop a specific pipeline for this multi-modalities study.
Findings
Radiotherapy leads to a significant decreasing of FA values across ROIs, voxels and individuals; the correlations of FA values is positively correlated with age/birth date, indicating young people have the higher risk to FA changes caused by radiotherapy; the "Inferior cerebellar peduncle left and right" are significant ROIs of the correlation between decreasing of FA values and radiotherapy. FA changes in a large majority of voxels of the "Inferior cerebellar peduncle left and right" ROIs have significant negative correlation across 202 individuals with the dose intensity; a large FA change leads to a reduced ( pvalue< 0.1 ) survival time.
References
This project has been submitted to OHBM 2017.Coauthor: Yan Jin, Fuller Clifton, Mohamed Abdallah Sherif Radwan, Ding Yao, Hongtu Zhu.