The results are presented in Table 4. The correspondence rate between residents and staff radiologists increased as the year of stay of CTPA interpretations is higher for PGY-4 &5 residents (0.90; 95% CI 0.82-0.98) compared to PGY-2 residents (0.72; 95% CI 0.63-0.80). For CTV interpretations, the concordance between residents and radiologists was less consistent, with a lower kappa for PGY-4 &5 residents (0.64; 95% CI 0.38-0.90) compared to PGY 3 residents (0.71; 95% CI 0.56-0.86). Table 2 shows the cross-classification of sites (right, left or both sides) for the 112 PE values that were interpreted as positive by both residents and radiologists. The overall agree rate was 92.0%. The agreements for “right”, “left” and “bilateral” were 88.0%, 93.0% and 92.0% respectively. There were indices of marginal homogeneity (χ2 (2 df) = 1.0, p = 0.61) with a kappa of 0.86 (95% CI 0.78-0.95). As shown in Table 4, there is a positive correlation between all values assessed by WB-CT and wb-MRI, which is statistically significant in the lumbar spine, long bones, ribs and other bones and the entire skeleton. The correlation is highest for the lumbar spine (Spearmans Rho = 0.81, p < 0.01). The correlation is weakest for the thoracic spine and statistically insignificant (Spearmans Rho = 0.25, p ̧ = 0.56).
There is a moderate correlation for the entire skeleton (Spearmans Rho = 0.61) and for other regions of the body (Spearmans Rho = 0.33 to 0.61). Non-MR experts have sufficient sensitivity when reading MRI in patients suspected of appendicitis, with good correspondence with mr experts, but the accuracy of the MR experts` reading was higher. The CTPA and CTV interpretations were classified as “positive”, “negative” or “undetermined”, based on preliminary resident reports and final radiology reports in concurrent cases. The location of PE has been saved. The study is the first to compare the correspondence between WB and WB-CT MRI. Greater interobserver concordance of WB MRI versus WB-CT has been demonstrated across the entire skeleton and for different body regions by experienced and young radiologists. Although higher values were derived from WB MRI compared to WB-CT for the detection of disease in long bones, there was no difference in the overall observer score. However, the study was limited by the inclusion of patients with confirmed myeloma and mean bone infiltration.
The study therefore did not take into account a significant number of patients with a lower disease burden, for whom MRI could be beneficial. Of the 696 CTPA assessed by residents, 128 (18.4%) were positive, 486 (70%) were negative and 82 (12%) were undetermined. On the other hand, radiologists reported 694 CTPA, of which 126 (18%) were positive, 493 (71%) negative and 75 (11%). Table 1 presents the cross-classification of CTPA assessment results for patients with interpretations available by both thinkers and staff radiologists. Among the 694 CTPA read by both residents and radiologists, the total compliance rate was 91.4% (634 out of 694), while there were a total of 60 divergent interpretations between residents and staff radiologists, a gap rate of 8.6%.