S. D. Pye

Single Microbubble Response Using Pulse Sequences: Initial Results

The study of acoustic scattering by single microbubbles has the potential to offer improved signal processing techniques. A microacoustic system that employs a hydrodynamically-focused flow was used to detect radiofrequency (RF) backscatter from single microbubbles. RF data were collected using a commercial scanner. Results are presented for two agents, namely Definity (Lantheus Medical Imaging, N. Billerica, MA, USA) and biSphere (Point Biomedical Corp, San Carlos, CA, USA). The agents were insonified with amplitude-modulated pulses, and it was observed in both agents that a subpopulation of microbubbles did not produce a measurable echo from the first-half amplitude pulse, but did produce a response from the full amplitude pulse and from a subsequent half amplitude pulse. The number of microbubbles in this subpopulation was seen to increase with increasing transmit amplitude. These results do not bear out the simple theory of microbubble-pulse sequence interaction and invite a reassessment of signal processing approaches.

P1F-6 Development of a Novel Experimental Set-Up to Allow Investigation of the Ultrasonic Backscatter from Microbubble Contrast Agents Attached to Surfaces

Developing applications of ultrasound contrast agents include targeting to areas such as inflamed plaque, drug delivery and gene therapy. In order to develop these techniques to their full potential the interaction of individual attached microbubbles with ultrasound and the mechanisms of transfer of material from microbubble to targeted areas needs to be fully understood. Aim: To develop an experimental set-up and technique suitable for determining the ultrasonic backscatter from individual, attached microbubbles and for the study of the transfer of material from single bubbles. A tank allowing acoustic and optical imaging was constructed with dimensions of: optical imaging section 34 times 16 times 4 cm, acoustic imaging section 30 times 16 times 30 cm. A microscope slide incorporated into the base of the optical imaging section maximised the quality of the microscope images. The imaging modalities were separate in the tank to allow precise characterisation of the acoustic field and to minimise reflections from surfaces other than the membrane. A holder comprising two Perspex rings held a 12 mum polyester membrane. A sliding device allowed movement of the membrane holder between the ultrasound field and the microscope. A membrane hydrophone with active element of 0.5 mm was used to determine the acoustic field at the surface of the membrane and within the tank. Copper (Cu) spheres of diameter 40-80 mum, attached to the membrane using poly-L-lysine (PLL), were used to assess the experimental set-up and to aid with alignment of the microscope optics with the acoustic field. A Sonos5500 scanner with S3 transducer of frequency range 1.26-3.75 MHz was used. The transducer was placed at an angle to the membrane to minimise the received echo, the area of interest on the membrane was 7.5 cm from the transducer. A sequence of 6 pulses was used. Previously the minimum detectable pressure with the ultrasound settings was determined to be 0.1 Pa. Cu spheres were used to assess the suitability of the set-up for studying microbubbles. Commercially available Definity microbubbles, mean diameter 1.1-3.3 mum, have been attached using PLL and imaged with 2 MHz ultrasound. The acoustic field was characterised for 1.26-3.75 MHz, for acoustic pressures of 300 and 550 kPa. Using the sliding device, Cu spheres attached to the membrane were found to reposition over the microscope objective within 20 mum of the original position which will allow optical imaging of contrast agents before and after insonation. The tank and set up has been shown to allow detection of the ultrasonic backscatter from individual particles attached to a membrane where the ultrasonic field at the location of the particle can be well calibrated. Definity attached to a membrane can be imaged and future experiments will investigate ultrasonic backscatter from attached single microbubbles for frequencies up to 11.1 MHz

Probing microbubble targeting with atomic force microscopy

Microbubble science is expanding beyond ultrasound imaging applications to biological targeting and drug/gene delivery. The characteristics of molecular targeting should be tested by a measurement system that can assess targeting efficacy and strength. Atomic force microscopy (AFM) is capable of piconewton force resolution, and is reported to measure the strength of single hydrogen bonds. An in-house targeted microbubble modified using the biotin-avidin chemistry and the CD31 antibody was used to probe cultures of Sk-Hep1 hepatic endothelial cells. We report that the targeted microbubbles provide a single distribution of adhesion forces with a median of 93pN. This interaction is assigned to the CD31 antibody-antigen unbinding event. Information on the distances between the interaction forces was obtained and could be important for future microbubble fabrication. In conclusion, the capability of single microbubbles to target cell lines was shown to be feasible with AFM.

The Edinburgh Pipe Phantom: characterising ultrasound scanners beyond 50 MHz

The ability to measure the imaging performance of pre-clinical and clinical ultrasound scanners is important but difficult to achieve objectively. The Edinburgh Pipe Phantom was originally developed to assess the technical performance of clinical scanners up to 15MHz. It comprises a series of anechoic cylinders with diameters 0.4 – 8mm embedded in agar-based tissue mimic. This design enables measurement of the characteristics (Resolution Integral R, Depth of Field LR, Characteristic Resolution DR) of grey-scale images with transducer centre frequencies from about 2.5 to 15MHz. We describe further development of the Edinburgh Pipe Phantom as a tool for characterising ultrasound scanners with centre frequencies up to at least 50MHz. This was achieved by moulding a series of anechoic pipe structures (diameters 0.045 – 1.5mm) into a block of agar-based tissue mimic. We report measurements of R, LR and DR for a series of 10 transducers (5 single element and 5 array transducers) designed for pre-clinical scanning, with centre frequencies in the range 15-55 MHz. Values of R ranged from 18-72 for single element transducers and 49-58 for linear array transducers. In conclusion, the pre-clinical pipe phantom was able to successfully determine the imaging characteristics of ultrasound probes up to 55MHz.

Characterising the performance of a high resolution ultrasound scanner for pre-clinical ultrasound imaging

Research using small animals continues to play a key role in biological, biomedical and veterinary science. In particular both mouse and rat models have become increasingly popular as research tools due to the fact that 90% of the genomic sequences in these rodents are identical to those found in humans. The versatility of employing a high resolution ultrasound scanner to perform small animal microimaging in vivo has recently been demonstrated [1]. However, in common with medical imaging systems, the technical performance of high resolution scanners is difficult to quantify. We have employed a novel measurement technique previously developed for medical imaging (the Resolution Integral) to study the grey-scale imaging performance of a high resolution scanner (Vevo 770, Visualsonics) with probe centre frequencies of 25-55 MHz. We designed and manufactured a high resolution test object containing 30 wall-less anechoic pipe structures in blocks of agar based tissue mimic. The pipe diameters ranged from 45 mum to 1.5 mm. Each probe was scanned over the surface of the test object and a series of images of each pipe was captured. The axial depth range over which each pipe could be visualised (L) was determined, and plotted as a function of plusmn, where plusmn is proportional to the reciprocal of pipe diameter. The Resolution Integral (R) was calculated for each probe by measuring the area under the curve. Characteristic Resolution and Depth of Field were also determined using the same set of measurements. Results were obtained using with probe models RMV710, RMV707B, RMV704, RMV708 and RMV711 (centre-frequencies 25, 30, 40, 55 and 55 MHz respectively). The measured values of R were 18, 23, 22, 21, 24 respectively, and the corresponding Depths of Field/Characteristic Resolutions were 5.4 mm/289 m; 5.3 mm/225 m; 3.0 mm/137 m; 2.8 mm/131m and 3.6 mm /145 m. We have successfully extended the range of application of Resolution Integral measurement to a high resolution microimaging system.

A numerical investigation into the effects of increasing pressure amplitude on microbubble resonance frequency

The analysis of the dynamical response of gas filled cavities surrounded by elastic capsules, acting as contrast agent microbubbles, is of importance in elucidating their full diagnostic use for ultrasound imaging. When the frequency of the external acoustic field is at or near the natural frequency of the bubble, resulting in resonance, larger amplitudes of oscillation than would otherwise occur are observed, enhancing the backscattered signal. It is therefore important to medical ultrasound imaging to be able to accurately determine the complex resonant behaviour of the bubble motion. In our work we consider the general Keller-Herring (K-H) model of Prosperetti and Lezzi. Appropriate variations in the cavity diameter, frequency and amplitude of the insonating field show computer-based simulations of microbubble response can be employed to determine resonance frequencies that are directly influenced by the driving pressure amplitude. The implication is therefore that the linear approximation to microbubble resonance is not applicable when considering insonating fields whose amplitudes exceed specific values which can be determined. A natural extension of the single bubble model is the investigation of interacting bubbles. It is shown that both pressure amplitude and the distance between bubble centres influence resonance frequency. A numerical investigation of the bifurcation properties of the K-H model is carried out: the driving frequency threshold for the onset of chaos decreases in the presence of an interacting microbubble or as the bubbles approach each other.

P5B-9 Investigation of the Response of Attached biSphere™ Microbubbles to Ultrasound

Knowledge of the behaviour of individual contrast microbubbles is essential if ultrasound contrast agents are to be developed to their full potential. In order to investigate the acoustic response of attached microbubbles an experimental system has been developed that can be used for sequential optical and acoustical imaging of attached particles. The aim of the work described here was to determine the feasibility of using the system for the study of single attached microbubbles. The experimental system, comprising tank and sliding device was mounted on a Leica inverted microscope alongside a Philips Sonos 5500 scanner and S3 transducer. The beam axis was positioned at 80deg to the surface of a 12 mum thick polyester membrane. Single copper (Cu) spheres were attached to the membrane and used to align the acoustic and optical fields. BiSpheretrade microbubbles (Point Biomedical) were treated to fluoresce and attached to the membrane using poly-L-lysine and an inversion technique. On one section of the membrane a patch (~1 mm diameter) of densely packed bubbles was used to confirm the alignment. In other areas on the membrane a diluted solution of biSpheretrade was attached which allowed one microbubble per field of view to be isolated. At times10 magnification the field of view was approximately 2 mm, with the ultrasound beamwidth at the area of interest being 4 mm. In order to determine the response of attached biSpheretrade, a range of acoustic pressures upto 1000 kPa and frequencies of 1.57 and 3 MHz were used to insonate patches and single bubbles of biSpheretrade. A 6 cycle pulse was used, and unprocessed backscattered RF data was captured. For patches of biSpheretrade, the backscatter signal was seen to remain constant with time at low acoustic pressures (400 kPa) while at acoustic pressures up to 1000 kPa an acoustic signal was observed that was consistent with gas escaping from attached bubbles. The fluorescent shells remained attached to the membrane after gas escape. For single attached biSpheretrade bubbles, no acoustic signal was detected at low acoustic pressures and at high acoustic pressures an acoustic signal from the escaping gas was recorded. It will be possible to modify the experimental system to allow the study of the transfer of materials from microbubbles to cells.

Attenuation Coefficients of the Individual Components of the International Electrotechnical Commission Agar Tissue-Mimicking Material

Tissue-mimicking materials (TMMs) are widely used in quality assurance (QA) phantoms to assess the performance of ultrasound scanners. The International Electrotechnical Commission (IEC) defines the acoustic parameters of up to 10MHz. To manufacture a TMM that closely mimics the acoustical properties of small animal soft tissue at high frequencies, the acoustic properties of each of the individual component ingredients used in the IEC agar-TMM recipe need to be quantified. This study was aimed at evaluating whether the overall attenuation coefficient of the IEC agar-TMM is the linear sum of the attenuation coefficients of each of its ingredients. Eight batches of agar-based materials were manufactured with different combinations of ingredients from the IEC agar-TMM recipe. The percentage concentration of each ingredient used in the individual mixes was identical to that specified in the IEC recipe. The attenuation of each of these batches was measured over the ultrasound frequency range 12-50MHz, and the attenuation value of the agar component was subtracted from the attenuation values of the other batches. Batch attenuation values, representing the attenuation of individual components within the IEC agar-TMM, were then summated and yielded attenuation values that accurately reproduced the attenuation of the IEC agar-TMM. This information forms a valuable resource for the future development of TMMs with acoustic properties similar to those of soft tissue at high frequencies.

The acoustical properties of IEC agar-based tissue mimicking material over the frequency range 4.5MHz to 50MHz - a longitudinal study

In this study the acoustic properties of thin slices of the IEC agar-based TMM were assessed over the frequency range of 4.5MHz to 50MHz for a period of 1 year at 3 monthly intervals. To maximize the acoustical stability of the TMM it was both assessed and stored in the same fluids as used in its manufacture (water, glycerol and benzalkonium chloride). A high frequency ultrasound scanner Vevo 770® and a scanning acoustic macroscope (SAM) system were used to measure the acoustical properties. The SoS measured from the TMM slices was found to be 1544.1 ± 4.2 ms-1 and a polynomial fit to the attenuation data was calculated to be 0.4842f + 0.008373f2 (R2=0.99). Batch to batch variation in acoustic properties and repeatability was also assessed (maximum variation of 7.4ms-1 for the mean SoS and 1dB cm-1 for the mean attenuation). These results demonstrate that the IEC agar-based TMM has excellent broadband acoustical stability over a period of 1 year.

Comparison of the acoustic response of attached and unattached BiSphere™ microbubbles

Two systems that independently allow the investigation of the response of individual unattached and attached microbubbles have previously been described. Both offered methods of studying the acoustic response of single microbubbles in well defined acoustic fields. The aim of the work described here was to investigate the responses of single attached microbubbles for a range of acoustic pressures and to compare these to the backscatter from unattached single microbubbles subjected to the same acoustic fields. Single attached BiSphereTM (Point Biomedical) microbubbles were attached to polyester with poly-L-lysine. Individual attached microbubbles were insonated at 1.6 MHz for acoustic pressures ranging from 300 to 1000 kPa using a Sonos5500 (Philips Medical Systems) research ultrasound scanner. Each microbubble was aligned to 6 cycle pulse, M-mode ultrasound beams, and unprocessed backscattered RF data captured using proprietary hardware and software. The backscatter from these microbubbles was compared to that of single unattached microbubbles subjected to the same acoustic parameters, microbubbles were insonated several times to determine possible differences in rate of decrease of backscatter between attached and unattached microbubbles. In total over 100 single attached microbubbles have been insonated. At 550kPa an acoustic signal was detected for 20 % of the attached microbubbles and at 1000 kPa for 63%. At acoustic pressures of 300kPa no signal was detected. Mean RMS fundamental pressure from attached and unattached microbubbles insonated at 800 kPa was 9.7 Pa and 8.7 Pa respectively. The ratio between the first two backscattered pulses decreased with increasing pressure. However, for unattached microbubbles the magnitude of the ratio was less than that of attached (at 550kPa mean ratio attached: 0.92 + 0.1, unattached: 0.28 + 0.2). There was no significant difference in the peak amplitude of the backscattered signal for unattached and attached microbubbles. BiSphereTM microbubbles comprise an internal polymer shell with an albumin coating, resulting in a stiff shell. BiSphereTM microbubbles do not oscillate in the same manner as a softer shelled microbubble, but allow gas leakage which then performs free bubble oscillations. The results here agree with previous acoustic and optical microscopy measurements which show that a proportion of microbubbles will scatter and this number increases with acoustic pressure. The lack of difference in scatter between the unattached and attached microbubbles may be attributed to the free microbubble oscillation being in the vicinity of the stiff shell, which may provide the same motion damping to a wall. Second pulse exposure shows that the wall becomes important in the survival of the free gas. These high quality measurements can be further improved by incorporating microbubble sizing to increase the specificity of the comparisons between unattached and attached microbubbles.