Figure 4.

Results from Test 2. a: Example echo amplitudes for k_a = 2s⁻¹ B₁ scaling = 1.1; both compartments are shown, but only the total echo amplitude would be observed experimentally. b: T₂ spectrum from data in (a) obtained using nonnegative least square (NNLS). Inset plot shows shorter T₂ peak; ${\widehat{\mathrm{T}}}_{2,\mathrm{b}}$ = 20.0 ms, $\widehat{\mathrm{f}}$ = 0.133 (obtained from peak area, shaded). c, d: $\widehat{\mathrm{f}}$ and ${\widehat{\mathrm{T}}}_{2,\mathrm{b}}$ as functions of k_a and B₁ scaling; $\widehat{\mathrm{f}}$ is primarily a function of k_a. When k_a = 0, $\widehat{\mathrm{f}}$ is estimated correctly; otherwise, $\widehat{\mathrm{f}}$ tends to be systematically underestimated. ${\widehat{\mathrm{T}}}_{2,\mathrm{b}}$ is primarily a function of B₁ scaling. e, f: $\widehat{\mathrm{f}}$ and ${\widehat{\mathrm{T}}}_{2,\mathrm{b}}$ as functions of k_a and ${\delta }_{\mathrm{b}}$ ( $\widehat{\mathrm{f}}$ depends on ${\delta }_{\mathrm{b}}$).