Optical fibres have revolutionized clinical practice in the form of the optical endoscope, and are now providing the framework for an entirely new endoscopic paradigm: all-optical ultrasound. Here we describe a phonon probe, which offers a route towards cellular resolution acoustic imaging – while also offering new capabilities such as label-free contrast and parallel spectroscopy – for future in vivo diagnostics.
The optical endoscope is perhaps the most intuitive inspection tool at the disposal of a clinician. In its simplest form, it effectively recreates human eyesight. Since its original development nearly two centuries ago, it has remained one of the cornerstones of minimally invasive medical procedures. Central to its minimal invasiveness is the humble optical fibre: a flexible strand of glass the size of a human hair that conducts light. Early on it was realized that by bundling many optical fibres together, light scattered from different spatial positions of the imaging target could be transmitted through separate channels of the bundle and used to form an image at the user-end. That is, the bundle enables spatial resolution, a feat which can also be provided by a single fibre attached to scanning equipment.
Resolution determines the smallest feature that can be observed by an imaging system. For example when using an optical microscope for histopathology, selecting lenses with insufficient resolution (generally correlated with magnification) may result in an inability to observe fine details within the tissue specimen. Conventional microscopy maximizes resolution with the following two parameters: short optical wavelengths and high numerical aperture lenses. However, these come with the caveats that shorter, bluer, wavelengths of light deliver more harmful energy to biological samples; and high numerical aperture lenses are often impractical to use. Unsurprisingly, endoscopy utilizes these same principles for increasing resolution, in addition to optical fibre bundles with minimized fibre-to-fibre spacing.
The field of endoscopy in the 21st century has become increasingly focused on maximizing imaging resolution. One of the motivations is rather simple. Can the invasiveness of biopsies be reduced by performing in vivo histopathology? What’s more, the invasiveness of a biopsy is not just limited to the excision process. Multiple excision sites, long processing times, sample viability and reusability, and the need for repeat procedures are all relevant complications in tissue diagnostics where the aim is to obtain a diagnosis as accurately and rapidly as possible. If an optical endoscope can be developed such that it contains the resolution of a bench-top microscope, the clinical community is one step closer to achieving the goal of high throughput in vivo diagnostics through optical biopsies. Endoscope developers such as Pentax and Mauna Kea have paved the way for such technology by integrating miniature optical lenses, high pixel density optical fibre bundles, miniature scanning equipment, and narrowband light sources (e.g. lasers) in place of or alongside classical endoscopic payloads. The most clinically translated of these technologies is the narrowband technique: confocal laser endomicroscopy (CLE), which, in expert hands, enables in vivo imaging comparable to ex vivo histological microscopy. In addition to providing microscopic visualizations of histology, CLE offers the unique opportunity to observe dynamic cellular processes in natural environments.
Thus far CLE has proven that neoplastic tissue in the colon, esophagus, stomach, and pancreas can be differentiated from healthy tissue for conditions such as ulcerative colitis, Barrett’s esophagus, and gastric cancer with accuracies in the range of 90–99%+ . Furthermore, when coupled with conventional techniques such as chromoendoscopy for identifying suspicious tissue on the macroscopic scale, the diagnostic yield of CLE increases further which ultimately allows more accurately targeted real-biopsies while minimizing the total number of invasive procedures and related consequences such as scarring.
Central to CLE and other endomicroscopy techniques, is the practice of fluorescence imaging. Since biological cells and tissue scatter light very similarly to water, it is often difficult to discern cells and sub-cellular features from their surroundings; i.e. there exists poor optical contrast. To circumvent this, the tissue is stained with molecules and proteins that absorb and re-emit light (fluorescence), and therefore allow individual cells – and even sub-cellular components such as the nucleus and cell membrane – to illuminate themselves from the perspective of the camera. Although fluorescence imaging is invaluable to the life and clinical sciences (and industry standard fluorophores such as fluorescein are known to be physiologically safe), there is great interest in developing label-free imaging techniques. Transitioning to label-free imaging would eliminate the need to closely monitor:
- non-damaging concentrations of fluorophores;
- energy blue light (often used for fluorescence excitation);
- exposure times; and
- fluorophore bleaching.
Whereas the optical properties between biological tissue and liquid – or throughout the tissue itself – may change very little, the mechanical properties of the tissue may vary much more significantly. This is the underlying principle of acoustic imaging, e.g. ultrasonography, where sound waves scatter more strongly at interfaces in density and/or sound velocity compared with the scattering of light at the same interfaces. In this way, acoustics natively offers a route to high-contrast imaging in biological materials. However, when engineering an acoustic endoscope to achieve imaging resolution comparable to that of optical endomicroscopy the challenges become less trivial.
As with optical imaging, the resolution of acoustic imaging techniques is fundamentally limited by the wavelength of the generated sound waves. Shortening the acoustic wavelength to the equivalent wavelength of blue light (e.g. λ≈300 nm) requires ultrasonic sources (transducers) that are exceedingly small, and therefore fragile and inefficient. The field of opto-acoustics has provided a solution to these caveats. If an ultrasonic transducer can be optically driven (instead of electrically), the size of the transducer can be reduced to the nanometre scale without sacrificing the robustness or sensitivity of the device. This concept works through the photoacoustic effect: the transducer (a metallic film) absorbs a short pulse of laser light and energy is converted into heat within the transducer, which relaxes in the form of mechanical vibrations (ultrasound) that propagate into the surrounding environment. By utilizing transducers with thicknesses of <100 nm, extremely high frequencies of ultrasound can be generated; e.g. in the GHz–THz range of the acoustic spectrum (compared with the kHz–MHz frequencies used in ultrasonographic equipment). In addition to being able to generate high-frequency (short wavelength) ultrasound using light, detection of the vibrations can also be carried out by optical means. One increasingly popular way of doing this is through the effect of Brillouin scattering.
In short, Brillouin scattering is the result of a particle of light (a photon) colliding with a particle of sound (a phonon). During the collision, the phonon will either give or take energy from the photon, i.e. the photon will shift frequency (its colour). The amount by which the photon shifts frequency is dependent on the frequency of the phonon. Therefore, by experimentally observing the shift in optical frequency, the frequency of the acoustic wave can be detected directly. Ultimately this acoustic frequency (called the Brillouin frequency) will be determined by the mechanical properties of the medium in which the phonon is travelling. Consequently, microscopes that can detect this Brillouin frequency are effectively mapping the stiffness of a specimen in the form of the longitudinal modulus, which is conceptually similar to the Young’s or Bulk moduli (although not identical).
Opto-acoustic techniques, such as Brillouin scattering, have the potential to provide the best of both worlds: imaging contrast which derives from the acoustic portion of the scattering interaction, yet imaging resolution determined by the optical portion of the interaction. This combination is ideal for cell imaging which demands both high contrast and resolution, and has led to the burgeoning field of Brillouin microscopy. Brillouin microscopes have been increasingly used for exploring questions that are pathological in nature. Central to these investigations is the role of mechanical properties in cells and tissue, a property which can be quantified through Brillouin microscopy, and therefore presents an important opportunity for quantitative imaging and diagnostics towards reducing the need for clinician-subjectivity.
It has been shown that cellular tissue can undergo radical transformations in stiffness under the influence of certain diseases and treatments, which is in addition to the role of stiffness in regulating basic cellular processes such as cell migration and division; however, a persistent question is whether stiffness abnormalities are causal or symptomatic of the underlying pathology. Thus far, Brillouin microscopy has been used in histological and spectroscopic settings to demonstrate that stiffness can be used as a biomarker in screening for bacterial meningitis , acute coronary syndrome  and keratoconus . From the perspective of therapeutics, Brillouin microscopy has begun to reveal unique mechanical responses and processes in colorectal carcinoma treatment , spinal cord growth and repair , Acanthamoeba castellani infection treatment , and amyotrophic lateral sclerosis development and treatment . Recently, our research team at the University of Nottingham has taken the first crucial step towards translating Brillouin microscopy from microscope to endoscope . This step involves two important milestones: (i) demonstrating 2D–3D imaging capability, and (ii) using standard optical fibre products that are found in commercial endoscopes (Fig. 1). We have developed a phonon probe which actively pumps GHz frequency ultrasound into a microscopic specimen (Fig. 2), and detects simultaneously 3D spatial information and stiffness-related information through Brillouin scattering (Fig. 3). All this with lateral resolution of the order of approximately 1 μm (similar to CLE), and axial resolution (dimension along the length of the fibre) on the nanometre scale, a combination which is unprecedented for optical or acoustic endoscopes.
Moving forward, our team will look to develop this technology into a clinically viable in vivo endoscopic imaging technique for minimally invasive opto-acoustic biopsies. Since the probe offers sub-microscopic resolution and label-free operation, it is capable of accessing the cellular regime accessed by CLE and histopathology. However, uniquely, it will be capable of quantifying the stiffness of cellular tissue, a parameter – which Brillouin microscopy has shown – that could lead to a whole new sector of pathology and diagnostics.
This article is an introduction to new developments in imaging
that have been recently published in the authors’ paper “Phonon
imaging in 3D with a fibre probe” in Light Science Applications
[2021; 10(1): 91] which can be seen at: https://www.nature.com/