Penetration depth measurement of a 6 MeV electron beam in water by magnetic resonance imaging

We demonstrate magnetic resonance imaging (MRI) visualization of a 6 MeV electron beam in ferrous-doped water; a 25 mm penetration depth was measured. Time domain nuclear magnetic resonance was used to investigate the effect of generated free radicals on the free induction decay (FID) in nondoped water; no apparent effects to the FID were observed. We show that MRI visualization of charged particle beams used in medical applications will require exogenous agents to provide contrast enhancement.

Figure 1

A 6 MeV electron beam was sent down a 3 m drift tube to the center of a 1.5 T MRI magnet. The electron beam exited the drift tube through a 0.15 mm titanium end window and impacted upon a glass vial containing a 1 mM ${\mathrm{Fe}}^{2+}$ $\mathrm{pH}=1.5$ aqueous solution.

Figure 2

Time sequence of transverse MRI images showing effect of a 6 MeV electron beam on MRI contrast through a 25 mm diameter vial of 1 mM ${\mathrm{Fe}}^{2+}$. Electron beam trajectory is normal to imaging plane. Electron pulse duration was $16\mu \mathrm{s}$. Multislice MRI $\mathrm{\text{parameters}}=3$ of 8 is shown (20 mm from where beam entered vial), image matrix $128\times 32\mathrm{pixels}$, $\mathrm{\text{slice thickness}}=5\mathrm{mm}$, $T_E=45\mathrm{ms}$, $T_R=814\mathrm{ms}$, acquisition time $\mathrm{\text{per image}}=26\mathrm{s}$. (a) Prior to e-beam impact; (b) 65 s post e-beam impact; (c) 195 s post e-beam impact; (d) 455 s post e-beam impact; (e) 663 s post e-beam impact; (f) 793 s post e-beam impact.

Figure 3

Time sequence of longitudinal (along magnets field and electron beam axis) MRI images showing effect of a 6 MeV electron beam on MRI contrast through a 25 mm diameter vial of 1 mM ${\mathrm{Fe}}^{2+}$. Electron beam trajectory is parallel to the imaging plane. The slice is along the vial midline. Electron pulse duration was $16\mu \mathrm{s}$. Multislice MRI parameters: $\mathrm{\text{slice}}=4$ of 8 shown (along central axis of vial), image matrix $128\times 17\mathrm{pixels}$, $\mathrm{\text{slice thickness}}=5\mathrm{mm}$, $T_E=86\mathrm{ms}$, $T_R=1300\mathrm{ms}$, acquisition time $\mathrm{\text{per image}}=22\mathrm{s}$. (a) Prior to e-beam impact; (b) 33 s post e-beam impact; (c) 187 s post e-beam impact; (d) 319 s post e-beam impact; (e) 451 s post e-beam impact; (f) 627 s post e-beam impact.

Figure 4

Transverse images of an e-beam passing through a 25 mm diameter vial of 1 mM ${\mathrm{Fe}}^{2+}$. The e-beam trajectory is normal to the eight imaging planes. Note the e-beam effectively stops prior to the imaging plane 35 mm from e-beam entry into the vial. The horizontal artifact in midplane is due to residual transverse magnetization from previous spin excitations. The intensity for all imaging planes, except that in the bottleneck (56 mm) are roughly equivalent for preirradiation. The rf homogeneity/sensitivity in the bottleneck is less than for the other imaging planes because that imaging plane is near the end of the rf coil. Images where the e-beam is present show a darker background, presumably from a shorter $T_2$. Beam pulse duration was $16\mu \mathrm{s}$. Multislice MRI parameters: 8 slices, image matrix $128\times 32\mathrm{pixels}$, $\mathrm{\text{slice thickness}}=5\mathrm{mm}$, $T_E=45\mathrm{ms}$, $T_R=814\mathrm{ms}$, acquisition time $\mathrm{\text{per image}}=26\mathrm{s}$.

Figure 5

Line plot through MR image of cylindrical vial of 1 mM ${\mathrm{Fe}}^{2+}$ aqueous solution after passage of a 6 MeV electron beam; the penetration depth of 25 mm for the electron beam is apparent. Note there is a background level of 160 (arbitrary units) from the Fricke solution.

Figure 6

Bulk spin lattice relaxation rate of 1 mM ${\mathrm{Fe}}^{2+}$ aqueous solution as a function of 6 MeV electron beam on time.