| Autor: |
Vaidya, Suthirth, Rao, Vishal, Shivalingappa, Shivakumar Swamy |
| Předmět: |
|
| Zdroj: |
Journal of Cancer Research & Therapeutics; 2017 Supplement, Vol. 13, pS145-S146, 2p |
| Abstrakt: |
INTRODUCTION:There has been increased use of Positron emission tomography (PET) using [18F] Fluoro-2-deoxy-d-glucose (FDG), a glucose metabolism analog, for prediction of outcomes and guiding the precision medicine in tumor detection, staging, and treatment planning and surveillance post therapy in Oral cancers. Accumulating evidence of recent literature suggests that pre-treatment FDG uptake could be used as a prognostic factor for predicting treatment outcomes and is motivated by the fact that radiotracer uptake in the tumor is dependent on the characteristics of the microenvironment; quantitative analysis of FDG uptake is based on changes in the standardized uptake value (SUV). Recent development in Radiomics provides a promising objective way for tumor assessment, which uses computerized tools to extract a large number of image features that capture additional information not currently used in clinic that has prognostic value. In this study, we explored radiomic features for extracting reliable information from PET images of oral cancer patients. Leading to development of noninvasive biomarkers which could be incorporated into the clinical planning process to modify patients' treatment based on their predicted loco-regional failure risk. Materials and Methods: 44 patients with squamous cell carcinomas of the oral cavity staged using PET-CT and patients treated with surgery followed by chemo-radiotherapy were included. Manual segmentation was done by a certified radiation oncologist on the PET-CT and total of 79 radiomic features were extracted from the Gross tumor volume keeping threshold of SUV within the GTV at 40% of maximum value. Medical Records showed local recurrences in 29 of 44 patients with a median follow-up of 3 years. Spearman's correlation index was used to identify highly correlated radiomic features with local recurrences. Respective Area under Curve (AUC) was calculated .Predictive model matrix was fit using logistic regression method for the highest correlated Biomarkers with local recurrences (AUC -0.85 and above). Results: All 44 patients were considered for analysis with no exclusions, two radiomic features V90 which is an intensity based histogram feature (AUC 0.95) and Extent which is a shape based feature (0.85 ) showed a strong co-relation with local recurrences in oral cancers (p. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
