Abstract
1784
Objectives The aim is to survey small lesion sizes detected in the latest clinical PET/CT which have been benefitted by technical improvements in the last decade. Rationale: Early detection may be the most effective method for cancer management due to the predominantly random nature of cancer formation [1-2]. PET/CT has been used in many countries for cancer screening. For early cancer detection, it is useful to know the capability of current PET/CT in detecting small lesions.
Methods The PET/CT images of 50 consecutive cancer patients (pre-treatment) from each of the three different clinical PET/CT were studied retrospectively. The PET/CT used were the Siemens Biograph, the GE TF960 and the United Imaging PET/CT (China). All are time-of-flight PET using L(Y)SO detectors. The patients from different systems were matched by cancer types, injection dose, delay times, patient weights, and imaging duration. The sizes of the lesions detected were measured following our tumor-analysis protocol using the PERCIST guideline with the patient's normal liver as a reference. The SUV thresholds used to measure lesion size were 2.1-3.6. Lesion size is defined as the largest lesion dimension on the PET image. Tumor size, MTV volume, SUVmean and SUVmax were extracted.
Results The results from the three PET/CT were merged. The smallest lesions detected were 4-5 mm with average SUVmean and SUVmax of 3.1 and 4.0 respectively. The smallest lesion MTV volume detected was in the range of 0.13 to 0.15 ml.
Conclusions Current state-of-the-art clinical PET/CT with technical improvements including time-of flight and detector-design improvements are capable of detecting small lesions for early cancer diagnosis applications. This preliminary finding will be further refined/confirmed from our ongoing analysis to include more retrospective patient studies.
Research Support This study is partly supported by NIH R01EB001038-11 grant and the Shanghai United Imaging Research fund.