Fig. 1From: Development of an algorithm for evaluating the impact of measurement variability on response categorization in oncology trialsBland-Altman plot and limits of agreement for the transformation candidates. a Intra-reader plot at baseline. The differences between the first and second measurements in relation to the average of the two measurements made by one reader of the tumor burden. b Intra-reader plot at post-treatment. The differences between the first and second measurements in relation to the average of the two measurements made by one reader of the tumor burden. c Inter-reader plot at baseline. The differences between the measurements in relation to the average of the measurements made by two readers of the tumor burden. d Inter-reader plot at post-treatment. The differences between the measurements in relation to the average of the measurements made by two readers of the tumor burden. Dotted lines: limits of agreement (LOA) from the model using the original lesion size; solid lines: LOA from the model using the log-transformation of the lesion size; dot-dashed lines: LOA from the model using cube root-transformation of the lesion size; dashed lines: LOA from the model using square root-transformation of the lesion sizeBack to article page