Search

Purpose: Instrumented ultrasonic tracking provides needle localisation during ultrasound-guided minimally invasive percutaneous procedures. Here, a post-processing framework based on a convolutional neural network (CNN) is proposed to improve the spatial resolution of ultrasonic tracking images. Methods: The custom ultrasonic tracking system comprised a needle with an integrated fibre-optic ultrasound (US) transmitter and a clinical US probe for receiving those transmissions and for acquiring B-mode US images. For post-processing of tracking images reconstructed from the received fibre-optic US transmissions, a recently-developed framework based on ResNet architecture, trained with a purely synthetic dataset, was employed. A preliminary evaluation of this framework was performed with data acquired from needle insertions in the heart of a fetal sheep in vivo. The axial and lateral spatial resolution of the tracking images were used as performance metrics of the trained network. Results: Application of the CNN yielded improvements in the spatial resolution of the tracking images. In three needle insertions, in which the tip depth ranged from 23.9 to 38.4 mm, the lateral resolution improved from 2.11 to 1.58 mm, and the axial resolution improved from 1.29 to 0.46 mm. Conclusion: The results provide strong indications of the potential of CNNs to improve the spatial resolution of ultrasonic tracking images and thereby to increase the accuracy of needle tip localisation. These improvements could have broad applicability and impact across multiple clinical fields, which could lead to improvements in procedural efficiency and reductions in risk of complications.

Optical ultrasound, where ultrasound is both generated and received using light, can be integrated in very small diameter instruments making it ideally suited to minimally invasive interventions. One-dimensional information can be obtained using a single pair of optical fibres comprising of a source and detector but this can be difficult to interpret clinically. In this paper, we present a robotic-assisted scanning solution where a concentric tube robot manipulates an optical ultrasound probe along a consistent trajectory. A torque coil is utilized as a buffer between the curved nitinol tube and the probe to prevent torsion on the probe and maintain the axial orientation of the probe while the tube is rotating. The design and control of the scanning mechanism are presented along with the integration of the mechanism with a fibre-based imaging probe. Trajectory repeatability is assessed using electromagnetic tracking and a technique to calibrate the transformation between imaging and robot coordinates using a known model is presented. Finally, we show example images of 3D printed phantoms generated by collecting multiple OpUS A-scans within the same 3D scene to illustrate how robot-assisted scanning can expand the field of view.

Laser-generated focused ultrasound (LGFU) transducers used for ultrasound therapy commonly have large diameters (6–15 mm), but smaller lateral dimensions (

Instrumented ultrasonic tracking is used to improve needle localisation during ultrasound guidance of minimally-invasive percutaneous procedures. Here, it is implemented with transmitted ultrasound pulses from a clinical ultrasound imaging probe that are detected by a fibre-optic hydrophone integrated into a needle. The detected transmissions are then reconstructed to form the tracking image. Two challenges are considered with the current implementation of ultrasonic tracking. First, tracking transmissions are interleaved with the acquisition of B-mode images and thus, the effective B-mode frame rate is reduced. Second, it is challenging to achieve an accurate localisation of the needle tip when the signal-to-noise ratio is low. To address these challenges, we present a framework based on a convolutional neural network (CNN) to maintain spatial resolution with fewer tracking transmissions and to enhance signal quality. A major component of the framework included the generation of realistic synthetic training data. The trained network was applied to unseen synthetic data and experimental in vivo tracking data. The performance of needle localisation was investigated when reconstruction was performed with fewer (up to eight-fold) tracking transmissions. CNN-based processing of conventional reconstructions showed that the axial and lateral spatial resolution could be improved even with an eight-fold reduction in tracking transmissions. The framework presented in this study will significantly improve the performance of ultrasonic tracking, leading to faster image acquisition rates and increased localisation accuracy.

Optical ultrasound imaging uses light to both generate and detect pulse-echo ultrasound. Recently, we presented a fibre-optic optical ultrasound imaging probe comprising 64 sources and a single receiver that allowed for video-rate, freehand imaging. However, its low number of sources limited the image contrast when using Delay-and-Sum reconstruction. Here, we present an alternative image formation paradigm for optical ultrasound based on model-based inversion, where the low number of sources allows for direct (i.e., non-iterative) inversion under modest hardware requirements. The model accurately incorporates the aperture geometry, frequency-dependent source directivity, and performance variation across the aperture, thereby reducing image artefacts associated with these properties. The method achieves a 15 dB gain in image contrast compared to Delay-and-Sum, at a similar image formation time.

Optical ultrasound is an emerging imaging paradigm that utilises optical fibres to emit and receive ultrasound. Their small lateral dimensions enable highly miniaturised imaging probes, for a range of interventional applications. This work presents a numerical study to determine the beam width, divergence angle and pressure retention for a range of probe geometries. Additionally, pulse-echo imaging of point scatterer targets is presented, and the impact of bandwidth filtering investigated. The results from this study provide the basis for developing a wide range of single sensor ultrasound devices.

Minimally invasive surgery offers significant benefits over open surgery in terms of patient recovery, complication rates, and cost. Accurate visualisation is key for successful interventions; however, no single imaging modality offers sufficient resolution, penetration, and soft-tissue contrast to adequately monitor interventional treatment. Consequently, multimodal interventional imaging is intensively investigated. All-optical ultrasound (AOUS) imaging is an emerging modality where light is used to both generate and detect ultrasound. Using fibre-optics, highly miniaturised imaging probes can be fabricated that yield high-quality pulse-echo images and are readily integrated into minimally invasive interventional instruments. In this work, we present the integration of a miniature (diameter: 800 µm), highly directional AOUS imaging probe into a commercially available white light urethroscope, and demonstrate the first real-time, 3D multimodal imaging combining AOUS and white light endoscopy. Through the addition of an electromagnetic tracker, the position and pose of the instrument could be continuously recorded. This facilitated accurate real-time registration of the imaging modalities, as well as freehand operation of the instrument. In addition, the freehand imaging paradigm allowed for “piece-wise” scanning where the instrument was retracted and repositioned without recalibration. The presented imaging probe and system could significantly improve the quality of image guidance during interventional surgery.

All-optical ultrasound transducers are well-suited for use in imaging during minimally invasive surgical procedures. This requires highly miniaturised and flexible devices. Here we present optical ultrasound transmitters for imaging applications based on modified optical fibre distal tips which allow for larger transmitter element sizes, whilst maintaining small diameter proximal optical fibre. Three optical ultrasound transmitter configurations were compared; a 400 µm core optical fibre, a 200 µm core optical fibre with a 400 µm core optical fibre distal tip, and a 200 µm core optical fibre with a 400 µm core capillary distal tip. All the transmitters used a polydimethylsiloxane-dye composite material for ultrasound generation. The material comprised a photostable infra-red absorbing dye to provide optical absorption for the ultrasound transduction. The generated ultrasound beam profile for the three transmitters was compared, demonstrating similar results, with lateral beam widths ca. 30 MHz. These highly flexible ultrasound transmitters can be readily incorporated into medical devices with small lateral dimensions.

Conventional ultrasound imaging probes typically comprise finite-sized arrays of periodically spaced transducer elements which, in the case of phased arrays, can result in severe grating and sidelobe artifacts. Whereas side lobes can be effectively suppressed through amplitude apodization (“AmpA”), grating lobes arising from periodicity in transducer placement can only be suppressed by decreasing the element pitch, which is technologically challenging and costly. In this work, we present source density apodization (“SDA”) as an alternative apodization scheme, where the spatial source density (and, hence, the element pitch) is varied across the imaging aperture. Using an all-optical ultrasound imaging setup capable of video-rate 2-D imaging as well as dynamic and arbitrary reconfiguration of the source array geometry, we show both numerically and experimentally how SDA and AmpA are equivalent for large numbers of sources. For low numbers of sources, SDA is shown to yield superior image quality as both side and grating lobes are effectively suppressed. In addition, we demonstrate how asymmetric SDA schemes can be used to locally and dynamically improve the image quality. Finally, we demonstrate how a nonsmoothly varying spatial source density (such as that obtained for randomized arrays or in the presence of source positioning uncertainty or inaccuracy) can yield severe image artifacts. The application of SDA can, thus, yield high image quality even for low channel counts, which can ultimately result in higher imaging frame rates using acquisition systems of reduced complexity.

n this study we report the integration of an all-optical ultrasound probe and robotic manipulator. The alloptical ultrasound probe comprised two optical fibres, a MWCNT/PDMS composite coated multimode fibre for ultrasound generation, and a concave Fabry-Perot fibre optic hydrophone for ultrasound reception. The ultrasound probe generated pressures in excess of 2 MPa at 1:5 mm, with a corresponding -6 dB bandwidth of ca. 30 MHz. The probe was built into a rigid endoscope (outer diameter: 5 mm, length: 300 mm) and mounted on a robotic manipulator. Ultrasound A-lines were acquired during robot manipulation in order to reconstruct a 3D image which was displayed as a point cloud. Large area images (80 × 80 mm) of a tissue mimicking gel wax phantom where acquired and displayed in real-time. This work demonstrates the potential for integrating miniature fibre optic ultrasound devices with robotics.

Fetoscopy is a technically challenging surgery, due to the dynamic environment and low diameter endoscopes often resulting in a limited field of view. In this paper, we report on the design and operation of a robotic multimodal endoscope with optical ultrasound and white light stereo camera. The manufacture and control of the endoscope is presented, along with large area (80 mm ×80 mm) surface visualisations of a placenta phantom using the optical ultrasound sensor. The repeatability of the surface visualisations was found to be 0. 446 ± 0.139 mm and 0. 267 ± 0.017 mm for a raster and spiral scan, respectively.

Radiofrequency ablation is widely used in cardiology as an effective minimally invasive treatment for atrial fibrillation. However, radiofrequency noise, electronic interference, low resolution and poor tissue contrast complicate real-time lesion monitoring using conventional imaging modalities such as magnetic resonance imaging or ultrasound imaging based on electronic transducers. Recently, a bench-top all-optical ultrasound imaging system, where ultrasound is both generated and detected using light, was presented (doi:10.1364/BOE.9.003481) that achieved high-resolution, video-rate 2D images. In this system, pulsed excitation light was focussed onto a nanocomposite membrane, where it was converted into ultrasound via the photoacoustic effect. Using scanning optics, the resulting optical ultrasound source was translated to synthesise a 1D source aperture comprising irregularly spaced ultrasound sources. Back-scattered ultrasound was detected using a single fibre-optic Fabry-Perot cavity. Here, this system (which is inherently insensitive to electromagnetic interference) was used to achieve the first video-rate, depth-resolved 2D images acquired during RF ablation using an all-optical ultrasound imaging setup. We used this system to monitor the formation of radiofrequency ablation lesions (max 30 W, 65°C, 60 s) in ex vivo chicken breast samples, at a frame rate of 9 Hz, resolution of 100 µm, an imaging depth >15 mm, and a contrast of up to 30 dB. With its high miniaturisation potential, all-optical ultrasound imaging shows great promise for guiding interventional procedures, where real-time ablation lesion visualisation could improve lesion delivery and patient outcome.

All-optical ultrasound imaging uses optical generation and detection of ultrasound to acquire pulse-echo images. Recent advances have resulted in efficient optical ultrasound sources, emitting pressures and bandwidths rivalling those generated by conventional electronic transducers. Two-dimensional imaging of biological tissues was achieved using a fibre-optic Fabry-Perot cavity and a nanocomposite generator membrane in which ultrasound was generated photoacoustically. Using scanning mirrors, excitation light was steered to consecutive locations, thus synthesising an acoustic source aperture with a geometry that could be arbitrarily and dynamically reconfigured. This unique capability of implementing different geometries on the same hardware allows for a direct comparison of the image quality obtained with different aperture geometries, which is difficult to achieve using conventional electronic transducers. Here we explore how the source aperture geometry affects the image quality through a set of numerical simulations and experiments. First, we determined that the image artefacts and corresponding contrast level depend strongly on the total number of A-scans (increasing from 200 to 1800 A-scans improved the contrast from 30 to 50 dB), irrespective of the number and locations of the detectors. Second, we demonstrated how parametric optimisation of the spatial optical ultrasound source distribution allowed for local (within a user-defined region of interest) or global image optimisation achieving an additional reduction in artefact level of up to 8 dB. Finally, we demonstrated video-rate, real-time 2D image acquisition using optimised source aperture geometries.

We propose and demonstrate a miniature microwave device capable of detecting wide-band ultrasound without the use of piezo materials. By the nature of the transduction mechanism, the scheme naturally implements orthogonal frequency division multiplexing (OFDM): each device can be assigned its own microwave band, mitigating electrical cross-talk between adjacent sensor elements for future imaging array systems. Furthermore, the entire array could be addressed with a single electrical connection, since addressing is in the frequency, rather than spatial, domain. We anticipate that the convenience afforded by this fully electronic platform will enable its use as a low cost and versatile solution in cases where high sensitivity is required over a large portion of the acoustic spectrum.

In this work, we demonstrate how the wide acoustic bandwidths generated by all-optical ultrasound imaging probes can be leveraged to dynamically trade-off image resolution with penetration depth. This dynamic trade-off was achieved through temporal modulation of the excitation light using bandwidth-limited linear chirps. The penetration depth of the all-optical ultrasound imaging probe used here could be arbitrarily set between 2 and 15 mm, to achieve inversely proportional spatial resolutions ranging from 128 − 500 μm (lateral)and 189 − 295 μn (axial). This added versatility in image parameters will enable seamless multi-scale ultrasound imaging that can strongly benefit interventional imaging.

Radio-frequency (RF)ablation is an effective minimally invasive treatment for atrial fibrillation, where RF fields are delivered by a catheter within the heart. In current clinical practice, this procedure is performed without real-time feedback about the viability of tissue beneath the ablation lesion. As a result, it can be challenging to perform complete ablation of the target tissue whilst avoiding damage to adjacent structures. Recently, all-optical ultrasound imaging, where ultrasound is both generated and detected using light, has been demonstrated as a viable imaging modality exhibiting high image resolution, strong miniaturisation potential, and resilience to interfererence by electromagnetic (EM)fields from the RF ablation catheter. In this work, the suitability of all-optical ultrasound to monitoring RF ablation lesion delivery is investigated using a bench-top imaging system. This system is capable of real-time, video-rate 2D imaging at an image depth of 15 mm, a resolution of 100 μm, a dynamic range of 35 dB, and a sustained frame rate of 9 Hz. In this paper we present the first real-time, video-rate all-optical ultrasound images acquired during RF ablation, and demonstrate how the imaging system was able to clearly visualise the lesion growth at a contrast of up to 30 dB. In addition, we demonstrated how lesion growth in inhomogeneous tissue can progress in an erratic fashion. High-quality imaging during treatment will enable optimisation of the ablation parameters during treatment, and will minimise both treatment times and complication rates.

All-optical ultrasound imaging, where ultrasound is generated and detected using light, has recently been demonstrated as a viable modality that is inherently insensitive to electromagnetic interference and exhibits wide bandwidths. High-quality 2D and 3D all-optical ultrasound images of tissues have previously been presented; however, to date, long acquisition times (ranging from minutes to hours) have hindered clinical application. Here, we present the first all-optical ultrasound imaging system capable of video-rate, real-time two-dimensional imaging of biological tissue. This was achieved using a spatially extended nano-composite optical ultrasound generator, a highly sensitive fibre-optic acoustic receiver, and eccentric illumination resulting in an acoustic source exhibiting optimal directivity. This source was scanned across a one-dimensional source aperture using a fast galvo mirror, thus enabling the dynamic synthesis of source arrays comprising spatially overlapping sources at non-uniform source separation distances. The resulting system achieved a sustained frame rate of 15 Hz, a dynamic range of 30 dB, a penetration depth of at least 6 mm, a resolution of 75 µm (axial) by 100 µm (lateral), and enabled the dynamics of a pulsating ex vivo carotid artery to be captured.

Percutaneous coronary interventions are widely performed minimally invasive procedures used to treat narrowing (stenosis) of arteries in the heart. Differential blood pressure measurements across a stenosis are invaluable to estimate the prognostic benefit of performing angioplasty and stenting via calculation of the fractional flow reserve. Achieving stable measurements from within pressure microcatheters and guidewires that are compatible with stenosed vessels, and which can be fabricated with low cost manufacturing methods, remains an important challenge. We have developed all-optical pressure and temperature sensors with a single optical fibre and sensing element. This approach provides simultaneous temperature and pressure measurements in a highly miniaturised device, with a simple construction method using low cost materials. Polymeric structures including membranes and domes are applied to the distal ends of single mode optical fibres. Temperature and pressure changes induce time-varying displacements of these structures, which are monitored using phase-resolved low-coherence interferometry. Phase measurements are acquired at 250 Hz with a sensitivity of approximately 0.2 rad/°C for temperature measurements between 20 and 45°C, and approximately 0.08 rad/mmHg for pressure between 760 and 1060 mmHg. In vivo studies in arteries and hearts of sheep and swine indicate that the sensors have sufficient sensitivity and speed for measurement of physiological pressure waveforms in clinical settings. We will discuss the integration of these sensors within medical devices, and the potential for providing additional physiological parameters with the same devices.

All-optical ultrasound imaging, where ultrasound is generated photoacoustically and back-scattered fields are detected with optically resonant structures, has distinct advantages over conventional piezoelectric or capacitive transducers. An absence of electrical connections facilitates probe miniaturisation and confers MRI compatibility, while similar or better sensitivities and bandwidths can be achieved. Previously reported all-optical systems used mechanical translation of a single source and receiver to scan a synthetic aperture, resulting in image acquisition times ranging from seconds to minutes. This frame-rate can be improved by reducing the number of A-scans per image; however, this increases the noise and artefact levels of the image. For conventional probes, typically employing unalterable periodic arrays, side- and grating lobes can give rise to substantial artefacts that are commonly suppressed using amplitude-weighted apodisation. In this work, however, the spatial density of the acoustic sources rather than their amplitudes is weighted to avoid array periodicity, in order to achieve a better image quality using the same aperture and number of A-scans.

An optical ultrasound generator was used to perform broadband (2–35 MHz) acoustical characterisation measurements of a nanocomposite comprising carbon nanotubes (CNT) and polydimethylsiloxane (PDMS), a composite that is commonly used as optical ultrasound generator. Samples consisting of either pure PDMS or CNT-loaded PDMS were characterised to determine the influence of CNTs on the speed of sound and power-law acoustic attenuation parameters. A small weight fraction (< 1.8%) of added CNTs was found to yield a prominent increase in the exponent of the power law, resulting in a significant increase in acoustic attenuation at higher frequencies. The speed of sound was found to be nearly identical, however. These results could prove useful in the numerical modelling and design of future optical ultrasound sources based on CNT-loaded PDMS.

Strongly directional ultrasound sources are desirable for many minimally invasive applications, as they enable high-quality imaging in the presence of positioning uncertainty. All-optical ultrasound is an emerging paradigm that exhibits high frequencies, large bandwidths, and a strong miniaturisation potential. Here, we report the design, modelling, and fabrication of a highly directional fibre-optic ultrasound transmitter that uses a composite of reduced graphene oxide and polydimethylsiloxane as the optical ultrasound generator. The ultrasound transmitter, which had an outer diameter of just 630 μm, generated ultrasound with a pressure exceeding 0.4 MPa for axial distances up to 16 mm, at a large bandwidth of 24.3 MHz. The ultrasound beam exhibited low divergence, with a beam diameter ranging between 0.6 and 2.1 mm for distances between 0 and 16 mm. The presented directional optical ultrasound source is hence well-suited to high-resolution interventional imaging.

Ultrasound arrays used for medical imaging consist of many elements placed closely together. Ideally, each element vibrates independently. However, because of mechanical coupling, crosstalk between neighboring elements may occur. To quantify the amount of crosstalk, the transducer velocity distribution should be measured. In this work, a method is presented to reconstruct the velocity distribution from far-field pressure field measurements acquired over an arbitrary surface. The distribution is retrieved from the measurements by solving an integral equation, derived from the Rayleigh integral of the first kind, using a conjugate gradient inversion scheme. This approach has the advantages that it allows for arbitrary transducer and pressure field measurement geometries, as well as the application of regularization techniques. Numerical experiments show that measuring the pressure field along a hemisphere enclosing the transducer yields significantly more accurate reconstructions than measuring along a parallel plane. In addition, it is shown that an increase in accuracy is achieved when the assumption is made that all points on the transducer surface vibrate in phase. Finally, the method has been tested on an actual transducer with an active element of 700 × 200 μm which operates at a center frequency of 12.2 MHz. For this transducer, the velocity distribution has been reconstructed accurately to within 50 μm precision from pressure measurements at a distance of 1.98 mm (=16λ0) using a 200-μm-diameter needle hydrophone.

Recent advances in optical transmission and reception of ultrasound have enabled pulse-echo imaging using fiber-optic probes that are suitable for guiding clinical intracardiac and intravascular procedures. In vivo images from pre-clinical models will be presented.

A miniature all-optical ultrasound imaging system is presented that generates three-dimensional images using a stationary, real acoustic source aperture. Discrete acoustic sources were sequentially addressed by scanning a focussed optical beam across the proximal end of a coherent fibre bundle; high-frequency ultrasound (156% fractional bandwidth centred around 13.5 MHz) was generated photoacoustically in the corresponding regions of an optically absorbing coating deposited at the distal end. Paired with a single fibre-optic ultrasound detector, the imaging probe (3.5 mm outer diameter) achieved high on-axis resolutions of 97 μm, 179 μm and 110 μm in the x, y and z directions, respectively. Furthermore, the optical scan pattern, and thus the acoustic source array geometry, was readily reconfigured. Implementing four different array geometries revealed a strong dependency of the image quality on the source location pattern. Thus, by employing optical technology, a miniature ultrasound probe was fabricated that allows for arbitrary source array geometries, which is suitable for three-dimensional endoscopic and laparoscopic imaging, as was demonstrated on ex vivo porcine cardiac tissue.

The fast nearfield method, when combined with time-space decomposition, is a rapid and accurate approach for calculating transient nearfield pressures generated by ultrasound transducers. However, the standard time-space decomposition approach is only applicable to certain analytical representations of the temporal transducer surface velocity that, when applied to the fast nearfield method, are expressed as a finite sum of products of separate temporal and spatial terms. To extend time-space decomposition such that accelerated transient field simulations are enabled in the nearfield for an arbitrary transducer surface velocity, a new transient simulation method, frequency-domain time-space decomposition (FDTSD), is derived. With this method, the temporal transducer surface velocity is transformed into the frequency domain, and then each complex-valued term is processed separately. Further improvements are achieved by spectral clipping, which reduces the number of terms and the computation time. Trade-offs between speed and accuracy are established for FDTSD calculations, and pressure fields obtained with the FDTSD method for a circular transducer are compared with those obtained with Field II and the impulse response method. The FDTSD approach, when combined with the fast nearfield method and spectral clipping, consistently achieves smaller errors in less time and requires less memory than Field II or the impulse response method.

In this research, we examined whether fixed pattern noise or more specifically Photo Response Non-Uniformity (PRNU) can be used to identify the source camera of heavily JPEG compressed digital photographs of resolution 640 x 480 pixels. We extracted PRNU patterns from both reference and questioned images using a two-dimensional Gaussian filter and compared these patterns by calculating the correlation coefficient between them. Both the closed and open-set problems were addressed, leading the problems in the closed set to high accuracies for 83% for single images and 100% for around 20 simultaneously identified questioned images. The correct source camera was chosen from a set of 38 cameras of four different types. For the open-set problem, decision levels were obtained for several numbers of simultaneously identified questioned images. The corresponding false rejection rates were unsatisfactory for single images but improved for simultaneous identification of multiple images.

In biomedical all-optical pulse-echo ultrasound systems, ultrasound is generated with the photoacoustic effect by illuminating an optically absorbing structure with a temporally modulated light source. Nanosecond range laser pulses are typically used, which can yield bandwidths exceeding 100 MHz. However, acoustical attenuation within tissue or nonuniformities in the detector or source power spectra result in energy loss at the affected frequencies and in a reduced overall system efficiency. In this work, a laser diode is used to generate linear and nonlinear chirp optical modulations that are extended to microsecond time scales, with bandwidths constrained to the system sensitivity. Compared to those obtained using a 2-ns pulsed laser, pulse-echo images of a phantom obtained using linear chirp excitation exhibit similar axial resolution (99 versus 92 $\upmu \text{m}$ , respectively) and signal-to-noise ratios (SNRs) (10.3 versus 9.6 dB). In addition, the axial point spread function (PSF) exhibits lower sidelobe levels in the case of chirp modulation. Using nonlinear (time-stretched) chirp excitations, where the nonlinearity is computed from measurements of the spectral sensitivity of the system, the power spectrum of the imaging system was flattened and its bandwidth broadened. Consequently, the PSF has a narrower axial extent and still lower sidelobe levels. Pulse-echo images acquired with time-stretched chirps as optical modulation have higher axial resolution (64 $\upmu \text{m}$ ) than those obtained with linear chirps, at the expense of a lower SNR (6.8 dB). Using a linear or time-stretched chirp, the conversion efficiency from optical power to acoustical pressure improved by a factor of 70 or 61, respectively, compared to that obtained with pulsed excitation.

As recently demonstrated, all-optical B-mode ultrasound imaging can be performed with a pair of coated optical fibres for transmission and reception that is translated to create a virtual array of elements. However, with translation in the in-plane dimension, the small lateral dimensions of the fibres results in out-of-plane acoustic divergence, which leads to image clutter and loss of sensitivity.

A novel fluidic actuation system has been developed for in situ imaging of anatomic tissues. The actuator consists of a micromachined superelastic tool guide driven by a pair of pneumatic artificial muscles. Two additional working channels allow easy interchange of instruments or sensing equipment. This paper describes the design and construction of the actuation system. Experimental results are also reported indicating a bending repeatability of 0.1 degrees and an operational bandwidth exceeding 8Hz. To show-case the performance of the device, the actuator was loaded with an all-optical ultrasound imaging probe. First scanned images of human placental tissue surface using an all-optical ultrasound probe are presented. While a model has been developed to estimate the probe position in space as function of the input pressure, in future work, this model will be complemented with additional sensor measurements of the bending probe taking into account the hysteretic behaviour of both muscles and nitinol structure. ispartof: pages:1415-1421 ispartof: 2015 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS) vol:2015-December pages:1415-1421 ispartof: 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems location:Hamburg date:28 Sep - 2 Oct 2015 status: published

Due to the difference between the absorption spectra of reduced and oxygenated haemoglobin, spectroscopic photoacoustic imaging can be used to determine, in vivo and non-invasively, the local blood oxygen saturation. Conventional methods achieve this by computing the absolute concentrations of reduced and oxygenated haemoglobin, which require models for, or computations of, the local fluence at depth. However, such estimates are potentially inaccurate or expensive to compute.

Photoacoustic imaging suffers from various sources of clutter to a greater extent than conventional ultrasound imaging. As a result, rather than a limited detector sensitivity or low fluence at depth, clutter is the main limitation in the maximum achievable image depth. In this work a method is presented that exploits the difference in spatial coherence across the transducer surface, expressed by the short-lag spatial coherence metric, to distinguish between signal and clutter. Using this metric as a weighting factor on a photoacoustic image, clutter is strongly suppressed while the photoacoustic amplitude is largely preserved. Using a clinical photoacoustic scanner, images were obtained of both a phantom and the first author's inner elbow, and strong clutter reduction was observed in both cases. Using the presented method, the effective imaging depth was doubled, both in phantoms and in vivo, without altering the fluence or the acquisition scheme.

Although molecularly targeted cancer therapies have shown great promise, it is now evident that responses are dependent upon the molecular genetic context. Spatial and temporal tumour heterogeneity renders biopsy of solid tumours unsuitable for determining the genetic profile of the disease, making adaptation of appropriate therapy difficult. We have utilized the tunable optical absorption characteristic of gold nanorods to assess the potential of photoacoustics for non-invasive multiplexed molecular imaging. Gold nanorods with resonance peaks at 700nm and 900nm were functionalised with in-house antibodies ICR55 and ICR62, targeted to HER2 and EGFR transmembrane receptors, respectively. Three human squamous carcinoma cell lines (LICR-LON-HN4 expressing high HER2 and low EGFR, LICR-LON-HN3 expressing intermediate levels of HER2 and EGFR and A431 expressing high EGFR and low HER2) were incubated with the targeted nanorods for 24 hours. Cells were then incorporated as simulated tumours in tissue-like phantoms composed of 7.5% gelatin containing 0.5% Intralipid® for optical scattering and imaged at a depth of 2.5 cm, using a new clinical in-house multi-spectral photoacoustic imaging system. Images were obtained from the cell inclusions for wavelengths ranging from 710 to 950 nm at 40 nm intervals, and the mean amplitude of the photoacoustic image was computed for each wavelength, to determine their relative receptor expression levels. The molecular profile of the cells obtained using multi-wavelength photoacoustics had substantial similarity to that obtained using flow cytometry. These preliminary results confirm selective uptake of the functionalised nanorods, which reflects the cellular expression of therapeutically important oncoproteins, and give an indication of the potential of photoacoustics for multiplexed molecular profiling.

A clinical spectroscopic eiphotoacoustic scanner is presented, evaluated and optimised. The scanner combines a clinical ultrasound scanner with commercially available lasers to generate photocoustic images that are inherently coregistered with the ultrasound images and can be acquired by freehand scanning. Arrays of graphite rods are used to determine the point spread function (PSF), signal-to-noise (SNR) and signal-to-clutter ratios (SCR) of the system and optimise the image quality by varying the illumination geometry. At a distance of 26 mm, mall spatial PSF extents are found (lateral 399 ± 80 μm, axial 340 ± 20 μm, elevational 1.43 ± 0.26 mm) and hence a high solution is obtained. Due to the strong optical scattering in biological tissue, the illumination geometry has virtually no effect on the image intensity, SNR and SCR, and hence practicality is the main concern when designing the clinical scan-head geometry.

Novel ultrasound arrays used in medical imaging contain many small elements (down to 100 μm) operating at high frequencies (upto tens of MHz). In such arrays, each element requires an individual electrical connection. Typically, these connections are made to soldering pads located some distance away from the transducer, which in turn are connected to the element by a pathway on a printed circuit board. However, these pathways are unshielded and therefore sensitive to electromagnetic interference and cross-talk. In this work, a casting-based fabrication process is presented where the room required for electrical connections is created vertically, i.e., in the direction perpendicular to the transducer surface. Using this process, chambers are created into which the cores of coaxial cables are inserted, and on top of these chambers the elements are built. As only the part of the cables inserted into the chambers is unshielded, the sensitivity to electromagnetic interference is strongly reduced. In addition, the elements share a single ground electrode which acts as a Faraday cage and further reduces interference. Using this fabrication process, a linear array was constructed containing eight elements of dimensions 100 μm by 350 μm operating at a center frequency of 21 MHz. The array is validated both acoustically and mechanically, and is shown to consist of eight virtually identical elements exhibiting high fractional bandwidths of more than 80%. Virtually no mechanical crosstalk is observed.

With intravascular ultrasound, high resolution cross-sectional images of arterial walls can be obtained. By pulling a catheter through the artery in the axial direction and stacking consecutive cross-sectional images, a three-dimensional image is obtained. However, cardiac and respiratory motion, combined with a high pull-back rate and a narrow axial radiation pattern, result in poor axial resolution and significant image distortion. To avoid these limitations, in this work the design of an axial linear transducer array is presented with which volumetric rather than cross-sectional images are obtained in every pullback position. The prototype consists of eight elements of dimensions 100 μm by 350 μm, with a kerf of 100 μm, operating at a center frequency of 21 MHz with a bandwidth of more than 80 %. For both an ex vivo bovine artery and a stent-emulating phantom, the image quality in the circumferential and radial directions is similar to that of a conventional catheter. However, in the axial direction a significant increase in image quality is achieved. With the array, it is possible to image both the struts of a stent phantom and the underlying structures, whereas with the conventional catheter the struts are not properly localised and the underlying tissue is not distinguishable from the stent.

Optimising the design of medical ultrasound transducers requires precise knowledge of the motion of the transmitting elements, especially if cross-talk between neighboring elements plays a significant role. Direct measurement of the motion of the elements is complicated. Therefore we propose to compute these velocity profiles from a pressure field measured in a plane parallel to the surface of the transducer. The method employs an inverse extrapolation of the measurements using an analytical deconvolution based on Weyl's representation of the Green's function. The method is tested for transducers operating in the MHz-range and with spatial dimensions in the order of 100 µm. Both measured and simulated pressure fields are used to investigate its performance. It is shown that relevant results, in terms of energy localisation and temporal behaviour, can be obtained for realistic distances between the transducer and measurement planes and for coarse sampling grids. In addition, the results show that sharp reconstructions of the velocity profiles are only possible if fine measurement grids are used and if the measurement distance is less than a wavelength of the ultrasound wave field.

In acoustic scattering simulations, unwanted reflections of the acoustic wavefield will occur when contrasts extend over the spatial boundaries of the numerical domain. One technique to suppress these reflections is to attenuate them in a locally, reflectionless absorbing boundary layer enclosing the spatial computational domain. This technique is commonly applied in FEM and FDTD simulations, and the layer is referred to as a Perfectly Matched Layer (PML). In this paper we will derive a PML formulation for the integral equation based acoustic scattering problem. The results of three-dimensional acoustic scattering simulations will be used to show its effectiveness.

Current intravascular ultrasound catheters generate high resolution cross-sectional images of arterial walls. However, the elevational resolution, in the direction of the catheter, is limited, introducing image distortion. To overcome this limitation, we designed and fabricated a linear array which can be rotated to image a three-dimensional volume at each pullback position. The array consists of eight rectangular piezo-electric elements of 350 µm by 100 µm operating at a center frequency of 21 MHz with a fractional bandwidth of 80 %, separated by a kerf of 100 µm. The array has been tested on both an ex vivo bovine artery and phantoms and, using the real aperture of the array, axially densely sampled images of the artery are obtained in every position. The array consistently yields significantly higher resolution in longitudinal images and more detail in radial images compared to a conventional catheter.