DIFFERENTIATING BENIGN FROM MALIGNANT LUNG LESIONS USING QUANTITATIVEPARAMETERS OF FDG PET IMAGES

Fluorine-18 labeled deoxyglucose positron-emission tomography (FDG-PET ) applications in oncology include the differential diagnosis of chest masses and single pulmonary nodules. However, FDG is not tumor-specif ic; rather, it also accumulates in inflammatory processes. This study was performed to identify image parameters that would improve the spec ificity of PET. Methods: Twenty-six patients who had benign and malign ant lung lesions were examined retrospectively. Positron-emission tomo graphy data were acquired in dynamic scanning mode after intravenous b olus of 250-402 MBq of FDG. Standardized uptake values (SUVs) were cal culated and Patlak analyses were performed in selected regions of inte rest in the PET images. Positron-emission tomography results were rela ted to histological diagnosis (N = 49) or clinical follow-up (N = 3). Results: The specificity and sensitivity of the original PET scan repo rts, which was based on visual image interpretation and loosely applie d SUVs, was 100% and 73%, respectively. Using the SUVs with a cut-off value of 3.8 and Kpat value with a cut-off at 0.025 min(-1) improved t he specificity to 81% and 85%. Conclusion: FDG-PET image interpretatio n can be facilitated by using SUV information or the accumulation rate of the radiotracer (Patlak). With additional validation, this method could have a significant cost-effective impact on the medical/surgical management of chest masses.