In vivo imaging using surface enhanced spatially offset raman spectroscopy (SESORS): balancing sampling frequency to improve overall image acquisition
Fay Nicolson,Bohdan Andreiuk,Eunah Lee,Bridget O’Donnell,Andrew Whitley,Nicole Riepl,Deborah L. Burkhart,Amy Cameron,Andrea Protti,Scott Rudder,Jiang Yang,Samuel Mabbott,Kevin M. Haigis
DOI: https://doi.org/10.1038/s44303-024-00011-9
IF: 17.694
2024-04-04
NanoScience and Technology
Abstract:In the field of optical imaging, the ability to image tumors at depth with high selectivity and specificity remains a challenge. Surface enhanced resonance Raman scattering (SERRS) nanoparticles (NPs) can be employed as image contrast agents to specifically target cells in vivo; however, this technique typically requires time-intensive point-by-point acquisition of Raman spectra. Here, we combine the use of "spatially offset Raman spectroscopy" (SORS) with that of SERRS in a technique known as "surface enhanced spatially offset resonance Raman spectroscopy" (SESORRS) to image deep-seated tumors in vivo. Additionally, by accounting for the laser spot size, we report an experimental approach for detecting both the bulk tumor, subsequent delineation of tumor margins at high speed, and the identification of a deeper secondary region of interest with fewer measurements than are typically applied. To enhance light collection efficiency, four modifications were made to a previously described custom-built SORS system. Specifically, the following parameters were increased: (i) the numerical aperture (NA) of the lens, from 0.2 to 0.34; (ii) the working distance of the probe, from 9 mm to 40 mm; (iii) the NA of the fiber, from 0.2 to 0.34; and (iv) the fiber diameter, from 100 μm to 400 μm. To calculate the sampling frequency, which refers to the number of data point spectra obtained for each image, we considered the laser spot size of the elliptical beam (6 × 4 mm). Using SERRS contrast agents, we performed in vivo SESORRS imaging on a GL261-Luc mouse model of glioblastoma at four distinct sampling frequencies: par-sampling frequency (12 data points collected), and over-frequency sampling by factors of 2 (35 data points collected), 5 (176 data points collected), and 10 (651 data points collected). In comparison to the previously reported SORS system, the modified SORS instrument showed a 300% improvement in signal-to-noise ratios (SNR). The results demonstrate the ability to acquire distinct Raman spectra from deep-seated glioblastomas in mice through the skull using a low power density (6.5 mW/mm 2 ) and 30-times shorter integration times than a previous report (0.5 s versus 15 s). The ability to map the whole head of the mouse and determine a specific region of interest using as few as 12 spectra (6 s total acquisition time) is achieved. Subsequent use of a higher sampling frequency demonstrates it is possible to delineate the tumor margins in the region of interest with greater certainty. In addition, SESORRS images indicate the emergence of a secondary tumor region deeper within the brain in agreement with MRI and H&E staining. In comparison to traditional Raman imaging approaches, this approach enables improvements in the detection of deep-seated tumors in vivo through depths of several millimeters due to improvements in SNR, spectral resolution, and depth acquisition. This approach offers an opportunity to navigate larger areas of tissues in shorter time frames than previously reported, identify regions of interest, and then image the same area with greater resolution using a higher sampling frequency. Moreover, using a SESORRS approach, we demonstrate that it is possible to detect secondary, deeper-seated lesions through the intact skull.