PET Enhanced CT Scan Performance in Cancer

The primary endpoint was the inter-observer reproducibility of the interpretation of the combined PET / enhanced CT scan (PET-CT) by anatomical region. Reproducibility was assessed for each of the 5 anatomical regions (thorax, abdomen, pelvis, bone, nervous system). Two independant pairs (B1 and B2), each composed of one nuclear physician and one radiologist interpreted the PET-CT examination and described each of the 5 anatomical régions according to 3 modalities (Presence of suspicious lesion(s); Presence of dubious lesion(s); Absence of suspicious and dubious lesion). The inter-observer reproducibility (inter-pairs of observers) was evaluated for each anatomical region by comparing the interpretations of the two pairs, using the weighted kappa concordance coefficient [ref = Fleiss J, Levin B, Cho Paik M. Statistical methods for rates and proportions. Third ed. 2003.].Interpretation by B1 after PET-CT examination (1 month after). Interpretation by B2 at the end of the study

The inter-observer reproducibility of combined PET-CT interpretations has been assessed globally for each patient. Same pairs of observer (B1 and B2) than for the primary endpoint evaluation interpreted the PET-CT examination in a global way and concluded for each patient. A weighted Kappa coefficient has been calculated from an identical methodology to that described for the primary endpoint evaluation. Interpretation by B1 was performed at least 1 month and 1 week after PET-CT examination. Interpretation by B2 was performed at the end of the study

For each of the 5 anatomical régions (thorax, abdomen, pelvis, bone, nervous system), we evaluated the reproducibility between the interpretations of the PET-CT by the nuclear physician alone (N1) and the independent pair (B2) composed by one nuclear physician and one radiologist . The nuclear physician alone (N1) and the independent pair (B2) interpreted the PET-CT examination independently and described each anatomical region.The inter-observer reproducibility has been evaluated for each anatomical region by comparing the interpretations of the nuclear physician alone and that one of independent pair of nuclear physician and radiologist, using the weighted kappa concordance coefficient [ref = Fleiss J, Levin B, Cho Paik M. Statistical methods for rates and proportions. Third ed. 2003.].Interpretation by nuclear physician alone (N1) was performed within 1 week of PET-CT examination. Interpretation by B2 was performed at the end of the study

The inter-observer reproducibility of combined PET-CT interpretations has been assessed globally for each patient. The nuclear physician alone (N1) and the independent pair (B2) interpreted the PET-CT examination independently in a global way and concluded for each patient. The inter-observer reproducibility has been evaluated at patient level by comparing the interpretations of the nuclear physician alone and that one of independent pair of nuclear physician and radiologist, using the weighted kappa concordance coefficient [ref = Fleiss J, Levin B, Cho Paik M. Statistical methods for rates and proportions. Third ed. 2003.]. Interpretation by nuclear physician alone (N1) was performed within 1 week of PET-CT examination. Interpretation by B2 was performed at the end of the study

For each anatomical region, the reproducibility of the injected CT scan was evaluated. The same radiologist evaluated the two injected CT scans (CT1 and CT2) and interpreted them (Presence of suspicious lesion(s) OR presence of dubious lesion(s) OR absence of suspicious and dubious lesion). Intra-observer reproducibility was analyzed by using the individual analysis by each radiologist. A weighted Kappa concordance coefficient was calculated per anatomical region using a methodology identical to that described for the evaluation of the proncipal endpoint. Interpretation of CT1 was performed befor inclusion. Interpretation of CT2 was performed at the end of the study.

The reproducibility of the injected CT scan was evaluated globally for each patient. The same radiologist evaluated the two injected CT scans (CT1 and CT2) and interpreted them (Presence of suspicious lesion(s) OR presence of dubious lesion(s) OR absence of suspicious and dubious lesion). Intra-observer reproducibility was analyzed by using the individual analysis by each radiologist. A weighted Kappa concordance coefficient was calculated using a methodology identical to that described for the evaluation of the proncipal endpoint. Interpretation of CT1 was performed befor inclusion. Interpretation of CT2 was performed at the end of the study.