All-optical ultrasound – rejuvenating an imaging workhorse

One of the most exciting recent developments in imaging consists of all-optical ultrasound. Unlike traditional ultrasound, which is achieved using piezoelectric transducers, optical systems perform ultrasonic generation via pulsed light. This is followed by optical reception of ultrasonic reflections from the tissue which is being imaged.

Reducing the need for trade-off
Though ultrasound is one of the most common medical imaging tools, conventional devices tend to be bulky and cannot typically be used at the same time as some other imaging technologies. This is why a hybrid combination of optics and ultrasound, coupled to inexpensive fibre-based probes for intravascular imaging, promises to open up new possibilities for medical imaging. 
In terms of imaging, optical techniques ensure satisfactory contrast levels, while ultrasound provides high resolution. Optical technologies can also be manipulated to generate low frequency ultrasound which yields greater penetration into tissue, or high frequency ultrasound to obtain higher resolution images, albeit at a shallower depth. In practical terms, such a combination also gives flexibility to physicians in how they use imaging technology to diagnose and treat medical problems. For example, to provide intravascular imaging and detection of conditions like plaque, ultrasound can provide details of morphology while the optical imaging highlights its composition.

Broadband imaging

In technical terms, traditional ultrasound image formation covers a narrow frequency band (usually half the central frequency), while signal generation in optical ultrasound is broadband (covering sub-MHz to several hundred MHz frequencies). In addition, the tomographic principles used by optical ultrasound generally entail data collection over wide angles. This improves image quality and resolution, while minimizing image artifacts.
Efforts to develop broadband all-optical ultrasound transducers date to over a decade. One prototype was developed and tested for high-resolution ultrasound imaging in 2007 at the University of Michigan, Ann Arbor. It consisted of a two-dimensional gold nanostructure on a glass substrate, followed by polydimethylsiloxane plus gold layers. The system achieved a signal-to-noise ratio of a pulse-echo signal of over 10dB in the far field of the transducer, where the centre frequency was 40MHz with −6dB bandwidth of 57MHz. In a paper published in the August 2008 issue of ‘IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control’, the developers of the system concluded that preliminary imaging results “strongly suggest that all-optical ultrasound transducers can be used to build high-frequency arrays for real-time high-resolution ultrasound imaging.”
Innovation not enough to offset drawbacks
As mentioned, the major driver of research into hybrid optical alternatives has consisted of limitations in conventional ultrasound systems. Such drawbacks persist in spite of developments in processing speed, a reduction in noise-to-signal ratios and enhancement in the quality and timing of image capture.
Although innovations such as matrix transducers have enabled the emergence of volumetric ultrasound and 3-D/4-D, elastography has offered physicians the ability to view both stiffer and softer areas inside of tissue.
Elastography uses b-mode ultrasound to measure the mechanical characteristics of tissues, which are then overlaid on the ultrasound image. However, its use in clinical practice remains complicated due to a wide range of techniques used by different manufacturers, alongside differences in parameters used to characterize tissues.
Other areas of innovation include micro-ultrasound, which harnesses ultrasound at microscopic levels and provides 3- to 4-fold improvements in resolution compared to conventional ultrasound. One of the first applications of micro-ultrasound is to allow better targeting of biopsies – for example, in the treatment of prostate cancer by urologists.

Ultrasound-modulated optical tomography
In the late 2000s, ultrasound-modulated optical tomography (UOT) showed considerable promise in imaging of biological soft tissues, with promising application in several areas, including cancer detection. UOT detects ultrasonically modulated light to localize and image subjects. The key limitation of UOT, however, is weak modulated signal strength.

Photoacoustic tomography

Considerable attention has also been given to photoacoustic tomography, which converts absorbed light energy into an acoustic signal. The technique provides compositional information on body tissue in real time without requiring any contrast agents. It also allows much higher depth penetration than conventional optical techniques. Photoacoustic tomography has been used for mapping the deposition of lipids within arterial walls.
Photoacoustic tomography begins by sending pulsed light into tissue, typically from a Q-switched Nd:YAG laser. This creates a slight hike in temperature which causes the tissue to expand, creating an acoustic response which is detected by an ultrasound transducer. The data is then used to visualize the tissue.
However, photoacoustic tomography systems have proved difficult to translate into clinical applications due to their high cost, as well as a relatively large footprint which requires a dedicated optical table to house the laser. On a technical level, moreover, a low pulse repetition rate (in the dozens of hertz) prevents photoacoustic tomography from being used in high frame rate imaging, which are required for clinical applications such as cardiac related problems, where the rate of blood flow is high, or in other similarly fast-moving settings.
There are nevertheless several efforts to cope with the challenges facing photoacoustic tomography. The first is enhancing signal-to-noise ratio and the depth of penetration of optical absorbers. Researchers at Purdue University in the US, who are at the forefront of investigations into the technique, believe that new optical manipulation techniques to maximize photon density might provide a way forward. They have recently announced development of a motorized photoacoustic holder, which allows manoeuvring the aim of the device and fine-tuning the depth to where light is focused. This, they believe, could significantly improve light penetration as well as the signal-to-noise ratio.
Other efforts seek to cope with fast-moving and dynamic settings. In Singapore’s Nanyang Technological University, for example, researchers have demonstrated up to 7,000 Hz photoacoustic imaging in B-mode, using a pulsed laser diode as an excitation source and a clinical ultrasound imaging system to capture and display the photoacoustic images.

All-optical ultrasound

All-optical ultrasound, which has recently catalysed maximum interest, involves using pulsed laser light to generate ultrasound. Scanning mirrors control where the waves are transmitted into tissue. After this, a fibre optic sensor receives the reflected waves, recombines them and creates a visualization of the area being imaged.

Bandwidth, acquisition time and electromagnetic interference

Such a modality exhibits wide bandwidth and satisfactorily addresses one of the major shortcomings of previous efforts at clinical application – namely, prolonged acquisition times (ranging from minutes to hours). Unlike conventional ultrasound imagers which use electronic transducer arrays to transmit sound waves into tissue and receive the reflections for reconstruction as images by a computer, all-optical ultrasound imagers are also immune to electromagnetic interference. As a result, an all-optical ultrasound system can be safely used alongside a magnetic resonance imaging (MRI) scanner, allowing physicians to obtain a more comprehensive picture of tissues around an area of interest, such as a tumour or blood vessel. Immunity from electromagnetic interference and MRI compatibility also means that all-optical ultrasound can be used during brain or fetal surgery, or for guiding epidural needles.

Miniaturization
The absence of electronic components gives yet another advantage, too. Components of conventional ultrasound devices are difficult to miniaturize for internal use. This is due to two factors: a drop in sensitivity after a reduction in the area of the active piezoelectric transducer, and the impact on size of the transducer due to the casing of the piezoelectric element and electrical insulation.
Miniaturization is particularly important in minimally invasive measurements such as medical endoscopy or for inspection of the lumen in non-destructive testing. Small-area detectors are also preferred in tomographic applications, given that detector size inversely correlates to spatial resolution.
Due to difficulties in miniaturization, most ultrasound devices use large, handheld probes placed against the skin. Although some high-resolution probes have been developed, they are considered too expensive for routine clinical use.
Unlike the above limitations for conventional ultrasound devices, the miniaturization of optical detectors (e.g. via interferometric resonators) does not impact on active detection area. In other words, there is no loss of sensitivity. Finally, optical components are not only easily miniaturized but also significantly less expensive to manufacture, compared to compacting electronic ultrasound systems.

First video-rate all-optical ultrasound system
The world’s first all-optical ultrasound system capable of video-rate, real-time imaging of biological tissue has been demonstrated by a research team from University College London (UCL) and Queen Mary University of London (QMUL). It was used to capture the dynamics of a pulsating ex-vivo carotid artery within the beating heart of a pig, and revealed key anatomical structures required to safely perform a transseptal crossing, namely left and right atrial walls, the right atrial appendage and the limbus fossae ovalis.
The researchers believe the new technology will allow ultrasound to be integrated into a wide range of minimally invasive devices in different clinical contexts, and provide ultrasound imaging of new and previously-inaccessible regions of the body. Above all, its real-time imaging capabilities allows differentiation between tissues at significant depths, helping to guide surgeons in some of the highest risk moments of procedures. This will reduce the chances of complications occurring in cases such as cardiac ablation.

Designed for clinical advantage

The new system from UCL and QMUL uses light guided by miniature optical fibres, encased within a customized clinical needle, which generate ultrasonic pulses. Reflections of these ultrasonic pulses from tissue are detected by a sensor on a second optical fibre, in order to provide the real-time imaging.
The developers based their design on a nano-composite optical ultrasound generator coupled to a fibre-optic acoustic receiver with extremely high sensitivity. In turn, harnessing eccentric illumination provided an acoustic source with optimal directivity. This was then scanned with a fast galvo mirror which provided video-rate image acquisition (compared to a time-frame of several hours in previous experiments). It also increased image quality in both 2D and 3D, and made it possible to acquire the images in different modes.
The scanning mirrors in the new system are flexible. They allow for seamless toggling between 2D and 3D imaging as well as a dynamically adjustable trade-off between image resolution and penetration depth. Unlike conventional ultrasound systems, these are achieved without requiring a swap of imaging probe. In a minimally invasive interventional setting, in particular, such probe swapping extends procedure times and introduces risks to the patient.
The technology has been designed upfront by the researchers for use in a clinical setting, with sufficient sensitivity to image moving tissue inside the body at centimetre-scale depth and fit into existing workflow. The researchers are currently working on developing a flexible imaging probe for free-hand operation, as well as miniaturized versions for endoscopic applications.