Room: Exhibit Hall
Purpose: [18F]FDG PET can provide glucose metabolism information at molecular level within the cancerous lesion, which is often heterogeneously distributed. This feature could reflect the basic biological characteristics of malignant tumor and their response to treatment. A simple, self-sufficient and easy-to-use mathematical model can be very helpful to utilize intra-tumor heterogeneity information extracted from [18F]FDG PET image data for treatment planning, prognosis and response assessment in routine cancer management. For this purpose, a new model has been developed and comparative study is reported in this abstract.
Methods: An heterogeneity index (H index) is defined to assess tumor heterogeneity by summing voxel-wise distribution of differential SUV from [18F]FDG PET image data. The summation is weighted by the distance of local SUV difference from the SUV-based tumor center and can thus yield increased values for tumors with peripheral sub-regions of high SUV which often serves as an indicator of augmented malignancy. The new model was compared with a widely-used model of Gray Level Co-occurrence Matrix (GLCM) for image texture characterization with phantoms and patient data of 3 lung cancer patients to evaluate its effectiveness and feasibility for clinical uses.
Results: The comparison of H index and GLCM parameters (Energy, Contrast, Local Homogeneity and Entropy) with phantoms demonstrate that H index is effective for characterization of the SUV heterogeneity in different phantom configurations. The value of H index increases along with the increasing heterogeneity level of SUV distribution in the phantoms. The single parameter of H index can quantify tumor heterogeneity in terms of the local SUV variation in patient images.
Conclusion: The new model has capacity to characterize the intra-tumor heterogeneity feature from [18F]FDG PET image data and the H index offers potential for clinical applications.