The mean of the RECIST measurements using constant thresholding was 15.59 mm with a negative bias percent error of , similar to the percent error for the -, -, and -dimensions. The standard deviation of the constant thresholding RECIST measurement was 0.09, and the CV was 0.60, also in line with the - and -dimension coefficients but much less than that of -dimension. RECIST measurements using adaptive thresholding showed lower levels of bias percent error and CV. For the volume measurement, the corresponding values were () and (4.94) for constant thresholding and () and (4.73) for adaptive thresholding. The bias percent error of the volume measurement was also negative and larger than that of the RECIST for both constant and adaptive thresholding because it reflects the bias in all three-dimensions (, and ). The standard deviation of the volume measurement was also much larger for the same reason; thus, the CV was useful for comparison purposes. The CVs of the -, -, and -dimension measurements of length with respect to the RECIST constant thresholding measure were 1.08, 1.17, and 6.21 times, respectively, while the volume measurement was 8.18 times that of the RECIST measure. The 9.46 CV ratio value relative to RECIST for adaptive thresholding was similar to constant thresholding. Observing that the degree of random error in the -dimension length and volume measurements as measured by the coefficient of variation are similar, the correlation coefficient was calculated and found to be 0.95 for both constant and adaptive thresholding.