Figure 4

Comparison of DC-MRF and ICP-MS concentration measurements. Tumor (a) Gd and (b) Dy concentration assessments are plotted against corresponding ICP-MS assessments. These plots include both single agent (n = 14 Gd-only, yellow triangle; n = 17 Dy-only, orange diamond), dual agent (n = 8, blue circle), and sham experiments (n = 1, black ‘x’). Inclusion of all experiments (n = 40) resulted in significant correlations between the DC-MRF and ICP-MS assessments (R² ≥ 0.83, p < 1e⁻⁶). Subgroup analysis of just the single agent and dual agent experiments also resulted in significant Pearson correlations for both Gd and Dy (R² ≥ 0.64, p < 0.007, Supplementary Fig. 3). All single agent experiments were included to assess the ability of the DC-MRF method to estimate concentration in the extreme case of one agent being absent from the sample. Note that this results in several “zero” points for each single agent experiment (i.e., Gd concentration is approximately zero for Dy-only experiments and Dy concentration is approximately zero for Gd-only experiments).