Omission errors in PET/CT

1154 Objectives: 1- Describe the epidemiology of PET misses of clinical relevance in PET/CT imaging. 2- Analyze the specific risk factors of missing lesions on PET imaging: related to hang-in protocol, review procedure, anatomical location, size and uptake intensity of the lesion, bio-distribution of the radio-pharmaceutical and cognitive bias of the interpreting physician. 3- Suggest means to reduce the risk of omission errors: use structured organ- adjusted review and reporting procedure, search for organs and patterns at-risk, use image segmentation software and alternative reconstruction algorithms to improve detection. Abstract: The large amount of imaging data presented for interpretation of PET/CT studies can lead to omitting clinically relevant lesions. In oncologic PET/CT, we found most commonly misses of bone metastasis, followed by incidental focal colon uptake and abnormal abdominal lymph nodes. Risk factors for missing lesions included overloaded hang-in protocol saturating the perception of the interpreting physician; non structured review and reporting, inappropriate windowing; non awareness to motion and mis-registration, vicinity of high-uptake physiological structures or other malignant lesions; overemphasis of CT features and cognitive faults related to ignorance of tracer distribution and negligence of peripheral body fields. To improve reading skills, we suggest structured reading and reporting habits: dedicated review of the coronal and sagittal planes for specific body areas; careful review of the maximum intensity projection (MIP) image at the end of the interpretation; use of CT-based bone segmentation on the PET images; reconstruction algorithm that increase signal-to-noise ratio on small findings; and awareness of the epidemiology of omissions.