The ¹⁹F concentration in a specific tissue can be quantified by direct comparison of the resulting image intensity with that of a reference sample. However, this approach relies on uniform spin excitation and signal reception. Due to tissue properties and MR coil characteristics, the local deterioration of the transmit (B₁⁺) and receive (B₁^-) fields between the targeted region of interest and the location of reference introduce errors in the quantified ¹⁹F signal. In previous work ¹, we applied a fast and accurate method for B₁⁺ and B₁^- fields mapping, based on the spoiled gradient echo (SPGR) approach ². This work shows the preclinical in vivo evaluation of the proposed ¹⁹F quantification technique with B₁⁺/B₁^--correction after systemic injection of hollow mesoporous silica sphere (HMSS) nanoparticles loaded with perfluoro-15-crown-5-ether (PFCE) (HMSS-PFCE) ³.

Considering B₁⁺ and B₁^- effects, the resulting measured signal intensity for SPGR acquisition with selected flip angle (FA) α_n is: $$I_{\alpha_{n}}(r)=cM_{0}(r)B_1^-(r)\frac{\sin(B_1^+(r)\cdot\alpha_{n})(1-E_{1})}{1-E_{1}\cos(B_1^+(r)\cdot\alpha_{n})}E_{2}, [1]$$ where M₀(r) is the equilibrium longitudinal magnetization, c a constant, $$$E_{1,2}=e^{-\frac{TR}{T_{1,2^{*}}}}$$$, B₁^-(r) the relative receive coil characteristic, and B₁⁺(r) the relative FA variation. For quantitative imaging, an image intensity I₀(r) proportional to M₀(r) is required. Neglecting c and E₂ in Eq. (1), I₀(r) can be approximated by: $$I_0(r) \approx \frac{I_{\alpha_{n}}(r)}{w_{ss}(r, \alpha_{n})\cdot B_1^-(r)}, [2]$$ where w_ss(r,α_n) is a steady-state factor depending on B₁⁺(r). Moreover, in case of using a transmit/receive coil B₁^-(r) = B₁⁺(r) allowing to compensate for transmit and receive effects after calculation of B₁⁺(r) from ratio of the signal intensities acquired with two different FAs.

The technique was tested in one NMRI mouse anesthetized using i.p. ketamine/xylazine injection to avoid any signal contamination from inhaled anesthesia gases. About 200µL of the HMSS-PFCE dispersion (theoretical load of 11.23mg PFCE) was injected systemically into the tail vein. A reference probe of 20mM (11.7mg/mL) concentration of PFCE diluted in n-Hexane was placed into the field-of-view (FOV). MR scanning was performed directly after injection (20 to 40min post-injection), 24h, and 96h post-injection. Two SPGR acquisitions with the following parameters: FOV = 3.2x4cm, matrix = 32x40, slice thickness 3mm, TE/TR = 1.7ms/350ms, NEX = 56, and FA = 30/45° resulting in 13min acquisition time per FA - could be performed within the anesthesia time.

¹⁹F signal was quantified based on the data acquired 24h post-injection. B₁⁺ mapping was performed on a slice-by-slice approach using previously measured PFCE T₁ value of 650ms. Pixel-wise concentration was calculated based on the non-corrected and B₁⁺/B₁^--corrected signal intensity images acquired at 30° and 45° respectively. Calculation of the total PFCE volume was done by integrating the ¹⁹F signal over all slices.

No fluorine signal from the mouse abdomen could be detected directly after intravenous injection of HMSS-PFCE nanoparticles. 24h post-injection pronounced fluorine signal could be detected exclusively in the mouse liver (Figure 1), which was completely gone 96h post-injection. Respective correction data considering transmit and receive effects could be calculated from 30° and 45° data. Resulting concentration maps derived from the signal intensity images acquired at 30° and 45° flip angles before and after correction are shown in Figure 2. An overview of the mean PFCE concentrations in the liver calculated for fluorine containing slices for corrected and non-corrected data is provided in Figure 3. Correction yielded similar ¹⁹F concentrations for different FAs, whereas without correction the resulting concentrations differed by up to 25%. The PFCE quantity in the liver as evaluated from the B₁⁺/B₁^--corrected concentration maps resulted to 9.73mg (for 30° FA image) and 9.59mg (for 45° FA image), corresponding to more than 85% of the injected amount of PFCE. Quantification from non-corrected data yielded high differences between the two FAs and lower absolute values of 7.10mg (for 30° FA image) and 8.23mg (for 45° FA image).Without correction, the inhomogeneous B₁⁺ profile yielded inaccurate quantification results as obvious by the huge differences of the ¹⁹F concentrations derived from two FAs. With the proposed correction, the resulting concentrations were almost independent on the FA indicating good correction of the transmit and receive effects. The PFCE amount in the liver calculated after B₁⁺/B₁^--correction (~10mg) does still not perfectly match the theoretical value of 11.23mg. Remaining differences may be attributed to actually lower amount of particles in injected solution, non-complete loading of the particles, variations of the particle properties and partly excreted or still circulating particles 24h after injection.In conclusion, proposed method for correction of B₁⁺ and B₁^- inhomogeneities enables accurate quantification of local ¹⁹F aggregations with acquisition times suitable for in vivo applications without requiring any modifications of the imaging sequence.