Watching while listening to the interaction of photons with bio-tissues
Optical coherence tomography (OCT) and photoacoustic microscopy (PAM) are two microscopic, three-dimensional, noninvasive imaging modalities that are based on different contrast mechanisms. OCT mainly images the optical scattering properties of a sample whereas PAM images optical absorption. Due to these different contrast mechanisms, OCT and PAM provide complementary information about biological tissues. OCT images the microanatomy of a sample (e.g., the histology-like cross-sectional image). It can also image the blood flow velocity by measuring the Doppler shift imparted to the probing light. Conversely, PAM maps the spatial distribution of light absorbers (e.g., hemoglobin and melanin) and is able to image the microvasculature and the associated blood oxygenation. As a result, a multimodal imaging system fusing OCT and PAM can provide more comprehensive information of biological tissues.1
The current strategy for fusing OCT and PAM is to use two different types of light sources for these two imaging modalities.2–6 OCT resolves the depth of a scatterer in biological tissues through coherence gating, which requires a broadband/low-coherence light source.7 PAM detects the laser-induced ultrasonic waves (photoacoustic waves) to form an image, which requires a pulsed laser, usually in the nanosecond range.8
It is natural to ask if a single light source could be used to accomplish OCT and PAM simultaneously. If so, the new technology would bring several advantages. First, it would produce images that are intrinsically registered in the lateral directions. Second, the two imaging modalities would be correlated, which would provide a unique opportunity for studying optical scattering and absorption in biological tissues. Finally, the method would minimize the risk of light damage by decreasing the total light exposure in multimodal imaging.
Our solution is to use a pulsed broadband light source for simultaneous PAM and OCT imaging. We recently demonstrated the feasibility of this concept,9 which we call optical coherence photoacoustic microscopy (OC-PAM): see Figure 1. In our experimental system, a neodymium-doped yttrium aluminum garnet (Nd:YAG) laser pumped a broadband dye laser (center wavelength 580nm, bandwidth 20nm), which was coupled into a Michelson interferometer to form the framework for OCT. In the sample arm of the interferometer, the light was scanned and focused onto the sample and underwent scattering and absorption. The back-scattered photons from the sample were combined with the light reflected in the reference arm and detected by a spectrometer, accomplishing OCT imaging in the spectral domain. In the meantime, the photoacoustic waves generated by the absorbed photons were detected by an unfocused needle ultrasonic transducer, accomplishing PAM imaging. With this detection scheme, each laser pulse can simultaneously generate an OCT A-scan (a one-dimensional scan along the path of the beam) and a PAM A-scan.
We imaged the ears of a Swiss Webster mouse in the experiment (see Figure 2). Simultaneously acquired OCT and PAM B-scan images (two-dimensional images made from a sequence of A-scans) are shown in Figure 2(a) and (b), respectively. Figure 2(c) shows the maximum-amplitude projection of the 3D PAM dataset. In the OCT image, different anatomical layers are resolved. The achievable imaging depth of the OCT mode in our OC-PAM system is about half the thickness achievable with standard near-infrared OCT due to the higher scattering coefficient of tissues in the visible spectrum. In the PAM images, only the signals generated from blood vessels can be observed. These imaging results demonstrated the complementary nature of the contrasts in the two imaging modalities.
One potential advantage of OC-PAM, related to using short-pulsed light for OCT, was not demonstrated in these experiments. In conventional continuous-wave spectral-domain OCT, any change of the sample (e.g., blood flow) during the exposure time of the camera (typically >10μs) cannot be revealed. Because the pulse width of OC-PAM is in the nanosecond range, it could allow much faster transient processes to be captured. In our current system, however, the high noise level of the light-source spectrum makes it impossible to carry out such observations.
In summary, we demonstrated the feasibility of using the same light source to accomplish OCT and PAM simultaneously. As a next step, we plan to use a more-stable pulsed broadband near-infrared light source to achieve better image quality and deeper imaging depth, and we will conduct quantitative analysis on the dataset. The technology could be invaluable for studying optical absorption and scattering in biological tissues.
This work is supported in part by National Institutes of Health grants NIH 7R21EB008800 to S. Jiao, NIH 1R01EY019951 to S. Jiao and H. F. Zhang, and NIH 1RC4EY021357 to H. F. Zhang, and by a Coulter Translational Award to S. Jiao.
Florida International University
Northwestern University