Imaging luminescent tattoo inks for direct visualization of linac and cobalt irradiation

Abstract

Purpose: Tattoo fiducials are commonly used in radiotherapy patient alignment, and recent studies have examined the use of UV-excited luminescent tattoo ink as a cosmetic substitute to make these visible under UV illumination. The goal of this study was to show how luminescent tattoo inks could be excited with MV radiation and imaged during beam delivery for direct visualization of field position.

Methods: A survey of nine UV-sensitive tattoo inks with various emission spectra were investigated using both UV and MV excitation. Images of liquid solutions were collected under MV excitation using an intensified-CMOS imager. Solid skin-simulating phantoms were imaged with both surface-painted ink and in situ tattooing during dose delivery by both a clinical linear accelerator and cobalt-60 source.

Results: The UV inks have peak fluorescence emission ranging from approximately 440 to 600 nm with lifetimes near 11-16 μs. The luminescence intensity is approximately 6x higher during the x-ray pulse than after the pulse, however, the signal-to-noise is only approximately twice as large. Spatial resolution for imaging was achieved at 1.6 mm accuracy in a skin test phantom. Optical filtering allows for continuous imaging using a cobalt source and provides a mechanism to discriminate ink colors using a monochromatic image sensor.

Conclusions: This study demonstrates how low-cost inks can be used as fiducial markers and imaged both using time-gated and continuous modes during MV dose delivery. Phantom studies demonstrate the potential application of real-time field verification. Further studies are required to understand if this technique could be used as a tool for radiation dosimetry.

Keywords: UV-sensitive ink; cherenkov imaging; fiducial marking; field verification; radiation therapy; scintillation imaging; tattoo ink.

Figure 1:

Normalized fluorescence emission of 9 UV-sensitive inks (A) and their corresponding normalized lifetime (B) based on 380nm excitation.

Figure 2:

Mean intensity of select UV-sensitive tattoo inks in ethanol solution, measured with 380nm UV excitation (A,) during the MV X-ray pulse (B) and after the pulse (C).

Figure 3:

Select UV-sensitive ink was painted on the surface of a PDMS-based skin-simulation phantom (E inset) and imaged during the X-ray pulse (A), after the pulse (B), and without time gating (C). The same phantom was imaged using a cobalt-60 source and 635 bandpass filter (D). The signal-to-noise (SNR) was measured for each ink using each imaging method (E), and was compared to the same configuration with a 600 longpass filter. The ratio of the SNR with (SNR_F) and without (SNR_NF) the filter provide insight on the noise (F). characteristics of each method. The SNR for each ink was measured using both a 550nm bandpass filter (left) and 635 bandpass filter (right, where the inset show the filter pass band (gray shaded) in relation to the ink emission (G).

Figure 4:

Red UV ink was tattooed into a skin-simulating phantom (top inset) to test system resolution under room-lights (top), time-gate to X-ray delivery (middle), and continuous acquisition during dose delivery (bottom).