Experiments using tissue-mimicking phantom were performed to demonstrate that MPA imaging is capable of detecting and identifying the regions labeled with dual-contrast nanoparticles with sufficient contrast and excellent contrast resolution.
shows the B-scan ultrasound image of the tissue-mimicking phantom where six inclusions were embedded in the phantom are marked. One of the inclusions (inclusion I) did not contain any contrast agent and was used as control, while other inclusions contained different types of contrast agents at different concentrations. The structure of the tissue-mimicking phantom and the locations of the inclusions were depicted in the B-scan ultrasound image. However, the US image cannot differentiate the inclusions due to the insignificant ultrasound contrast from the nanoparticles. In contrast, strong photoacoustic signals were detected only from the inclusions containing photoacoustic contrast agent: Au NRs (inclusion II) or dual-contrast liposomal nanoparticles (inclusions IV, V and VI). To quantitatively investigate the contrast enhancement in the PA image by nanoparticles, the magnitude of the averaged photoacoustic signal from each marked region was calculated and displayed in . The height of each column represents the magnitude of the averaged photoacoustic signal from the corresponding region, and the error bar shows the standard deviation of the photoacoustic signal. The PA image of the tissue-mimicking phantom is also shown in .
Fig. 5 (a) The ultrasound (US) image of the tissue-mimicking phantom with six inclusions. The background of the phantom was prepared by mixing gelatin with optical contrast agent to represent the endogenous chromophores in native tissue. The inclusions were (more ...)
As evident from , the inclusions that contain either Au NRs or liposomal nanoparticles have elevated optical absorption compared to the background. In addition, PA imaging is sensitive to the spatial variations of optical absorption; the differences in photoacoustic signals from the inclusions were consistent with the concentrations of the plasmonic nanoparticles. As shown in , the control inclusion (inclusion I) and the inclusion containing Fe3O4 NPs (inclusion II) did not generate significant photoacoustic signals. This was expected since light absorption at a wavelength 800 nm was insignificant. Inclusions II, IV and the left region of inclusion VI containing high concentration of Au NRs, either isolated or encapsulated in liposomal nanoparticles, generated intense photoacoustic signals. Inclusion IV and the right region of inclusion VI contained low concentrations of liposomal nanoparticles, which encapsulated fewer Au NRs, therefore produced weaker photoacoustic signals. Overall, the magnitude of photoacoustic signal is representative of concentration of plasmonic nanoparticles. However, there are also noticeable photoacoustic signals generated from the background material (mimicking endogenous chromophores in the native tissue). These background signals reduce the contrast of PA imaging and ability to identify the regions labeled by nanoparticles.
The MMUS image in , on the other hand, could identify the regions labeled with magnetic nanoparticles with high contrast because the tissues have fairly low magnetic susceptibility compared to magnetic nanoparticles. The graph shown in indicates the mean value and standard deviation of magneto-motive displacement measured within a rectangular window measuring 1.5 mm axially and laterally located at the center of each inclusion. The detected displacement in the background of the tissue-mimicking phantom was around 5 μm. The control inclusion (inclusion I) and the inclusion with Au NRs (inclusion II) exhibit almost no contrast in the MMUS image. Other inclusions containing Fe3O4 NPs or liposomal nanoparticles could be easily identified in the MMUS image. The largest magneto-motive displacement (around 35 μm) was measured in the inclusion containing Fe3O4 NPs (inclusion III). Although inclusion IV contained the same concentration of Fe3O4 NPs encapsulated in liposomes, the detected displacement from inclusion IV was around 30 μm. The slight difference in the displacements between inclusion III and IV is likely due to additional weight of Au NRs in liposomal nanoparticles. Comparing the displacements detected from inclusions IV and V, which contained different concentrations of liposomal nanoparticles, it is clear that as the concentration of the magnetic nanoparticles decreases, the induced displacement decreases accordingly. However, the map of magneto-motive displacement in inclusion VI was approximately uniform although the inclusion contained two regions (left and right) with different concentrations of liposomal nanoparticles. In fact the difference of concentrations of nanoparticles between left and right regions in inclusion VI was exactly the same as the difference between inclusion IV and inclusion V. Therefore, the MMUS imaging has limited ability to distinguish the regions containing spatially varying concentrations of magnetic nanoparticles if the regions are close to each other and mechanically connected. Thus, MMUS imaging exhibits low contrast resolution within the tissue with sub-regions containing different concentration of nanoparticles. However, the MPA image of the tissue-mimicking phantom shown in retains the best properties of each subsequent imaging technique. Specifically, the MPA image identifies the inclusions containing dual-contrast agent, liposomes encapsulating Fe3O4 NPs and Au NRs. Indeed, only the inclusions containing dual-contrast liposomal nanoparticles (inclusions IV, V and VI) were color-coded in the MPA image, and the signals from other inclusions and background were highly reduced. Furthermore, the regions with different concentrations of dual-contrast nanoparticles could be clearly distinguished using MPA imaging. Therefore, MPA imaging was capable of accurate representation of the tissue labeled with dual-contrast agents in the phantom.
The ex-vivo experiment with macrophages was designed to test the ability of MPA imaging to detect cells labeled with nanoparticles within the tissue. The US, PA, MMUS and MPA images of a tissue sample injected with nanoparticle-labeled macrophages are shown in
. The B-scan ultrasound image shown in visualizes the cross-sectional view of the tissue sample with injected macrophages. However, the macrophages cannot be easily identified because ultrasound backscattering from macrophages is similar to that of the background tissue, and the nanoparticles internalized by the cells do not provide significant contrast in ultrasound images. The PA image shown in could visualize the labeled cells in the tissue based on the optical absorption from the nanoparticles. But noticeable photoacoustic signals were also detected from background tissue, especially in the upper boundary of the tissue, because of the strong light absorption by endogenous chromophores within the tissue. Therefore, the contrast in the PA image is reduced. The MMUS image shown in identifies the presence of nanoparticles inside the tissue and suggests the location of nanoparticle-labeled macrophages with sufficiently high contrast. Indeed, the magneto-motive displacement in the region containing macrophages was around 100 µm, while the displacement from the background tissue was around 8 µm. Finally, the MPA image, obtained from the co-registered PA, MMUS, and US images and displayed in , identified the nanoparticle-labeled macrophages with sufficient contrast, excellent contrast resolution and high spatial resolution with the anatomic features of the imaged tissue.
(a) US image, (b) PA image, (c) MMUS image and (d) MPA image of ex vivo tissue sample injected with macrophage labeled with Au NRs and Fe3O4 NPs. The images cover area measuring 5.4 mm axially by 4.5 mm laterally.
Our results indicate the feasibility of MPA imaging to visualize the presence and location of nanoparticles inside tissue/cells with the spatial resolution of ultrasound imaging and enhanced contrast based on both the optical absorption and the magnetic susceptibility of the dual-contrast agent. Magneto-photo-acoustic images of the tissue-mimicking phantom and tissue sample were obtained by combining the co-registered US, PA and MMUS images. In the MPA image, only the signals from cells or tissue labeled with dual-contrast agent were selected and displayed over the anatomical content of the tissue, while the background or nonspecific signals were significantly suppressed. Therefore, MPA imaging enhances the contrast between the nanoparticle-labeled cells or tissues and the surrounding tissue. In the desired region labeled with dual-contrast nanoparticles, the MPA image provides high contrast resolution utilizing the sensitivity of photoacoustic signals to the variation of optical absorption caused by different concentration of nanoparticles. In addition, MPA imaging retains high spatial resolution determined by the ultrasound imaging system. Finally, the alignment of the data sets from different modalities in MPA imaging is simple and accurate because of the shared detection system. Therefore, the MPA image allows the improved spatial localization of the desired cells or tissue regions labeled with magneto-plasmonic nanoparticles with high sensitivity.
The liposomes containing Au NRs and Fe3
NPs were used as the dual-contrast agent in our experiments. The nanoconstructs were designed to exhibit both plasmonic resonance in near–infrared (NIR) spectral region and high magnetic susceptibility. The liposomal nanoparticles allowed flexibility in the loading ratio of Au NRs to Fe3
NPs. In addition to liposomal nanoparticles, various other hybrid nanoparticles could also be used as the contrast agent for MPA imaging [22
]. Furthermore, new particles optimized for MPA imaging can be designed [20
Magneto-photo-acoustic imaging is a promising tool for various biomedical applications. For instance, MPA imaging can assess mechanical and optical properties of soft tissue. Photoacoustic imaging can map optical absorption property of the tissue [3
], while the displacements measured in MMUS imaging can indicate the elasticity and viscosity of the soft tissue [12
]. Since there is significant correlation between diseases and local changes of soft tissue properties detected using MPA imaging, the MPA imaging technique has potential to detect the pathologies at early stages [6
]. Furthermore, the MPA imaging system can be used to guide and assess photothermal therapy by using the dual-contrast agent [13
]. Because of the presence of the magnetic component, the dual-contrast nanoparticles can be actively accumulated into the desired region using an external magnetic field. Based on the optical absorption property, the nanoparticles can lead to localized thermal damage by absorbing the radiant energy from the laser. The efficient targeting of nanoparticles through the magnetic field increases the effectiveness of the treatment and reduces the required dosage of photoabsorbers, thereby reducing the side effects associated with general systematic administration of nanoparticles. In addition, the dynamic MPA imaging of the targeted tissue can indicate the presence of the photo absorbers and assess the therapeutic outcome. The temperature maps measured with PA imaging [26
] and the tissue elasticity measured with MMUS imaging [12
] are important parameters expected to change significantly during the photothermal therapy.
Finally, molecular MPA imaging could be realized by functionalizing the surfaces of the dual-contrast nanoconstructs, allowing for improved spatial localization of the targeted cells in the context of the anatomic map of the tissue. Since the molecular localization of nanoparticles could be indicative of specific physiology, MPA imaging might provide a promising platform to noninvasively diagnose and characterize pathologies.