151751-Najiba-Chargi

37 Diagnostics: validation of skeletal muscle mass measurement IMAGE ANALYSIS Delineation of muscle tissue at the level of C3 was successful in all patients. Six patients (8.6%) had evident growth of a lymph node metastasis into the SCMmuscles. In these 6 patients, the SMA of the affected SCM muscle was substituted by the SMA of the unaffected, contralateral SCM muscle. CORRELATION BETWEEN SMA AT C3 AND L3 SMA at the level of C3 was not normally distributed (Shapiro-Wilk test <0.05). Spearman rank correlation analysis showed a good correlation between SMA at C3 and SMA at L3 (Spearman’s r s = 0.75; p < 0.01). Figure 2 shows the direct correlation between SMA measurements at the level of C3 and L3. Correlation between SMA at C3 and SMA at L3 was higher than the correlation between cross-sectional area of the paravertebral muscles only at C3 and SMA at L3 (Spearman’s r s  = 0.75 versus r s  = 0.70). Figure 2. Correlation between cross-sectional SMA (CSMA) at the level of C3 and (actual) cross-sectional SMA (CSMA) at the level of L3 SMA at L3 was estimated from SMA at C3 using the multivariate formula as described earlier (Formula 1). Actual SMA at L3 and estimated SMA at L3 were normally distributed (Shapiro-Wilk test: p > 0.05). Figure 3 shows the correlation between the estimated SMA at L3 and the actual SMA at L3. Pearson correlation analysis showed a high correlation between the estimated SMA at L3 and the actual SMA at L3 ( r = 0.82; p < 0.01). The mean difference between the estimated SMA at L3 and the actual SMA at L3 was calculated (mean -3.1 cm 2 , SD 5.9 cm 2 ), meaning that the estimated SMA at L3 was slightly lower than the actual SMA at L3. In 13 of 200 patients (7%) the estimated and actual SMA at L3 differed more than 1.96 standard deviation from the 2