A theranostic 3D ultrasound imaging system for high resolution image-guided therapy

Abstract

Microbubble contrast agents are a diagnostic tool with broad clinical impact and an increasing number of indications. Many therapeutic applications have also been identified. Yet, technologies for ultrasound guidance of microbubble-mediated therapy are limited. In particular, arrays that are capable of implementing and imaging microbubble-based therapy in three dimensions in real-time are lacking. We propose a system to perform and monitor microbubble-based therapy, capable of volumetric imaging over a large field-of-view. To propel the promise of the theranostic treatment strategies forward, we have designed and tested a unique array and system for 3D ultrasound guidance of microbubble-based therapeutic protocols based on the frequency, temporal and spatial requirements. Methods: Four 256-channel plane wave scanners (Verasonics, Inc, WA, USA) were combined to control a 1024-element planar array with 1.3 and 2.5 MHz therapeutic and imaging transmissions, respectively. A transducer aperture of ~40×15 mm was selected and Field II was applied to evaluate the point spread function. In vitro experiments were performed on commercial and custom phantoms to assess the spatial resolution, image contrast and microbubble-enhanced imaging capabilities. Results: We found that a 2D array configuration with 64 elements separated by λ-pitch in azimuth and 16 elements separated by 1.5λ-pitch in elevation ensured the required flexibility. This design, of 41.6 mm × 16 mm, thus provided both an extended field-of-view, up to 11 cm x 6 cm at 10 cm depth and steering of ±18° in azimuth and ±12° in elevation. At a depth of 16 cm, we achieved a volume imaging rate of 60 Hz, with a contrast ratio and resolution, respectively, of 19 dB, 0.8 mm at 3 cm and 20 dB and 2.1 mm at 12.5 cm. Conclusion: A single 2D array for both imaging and therapeutics, integrated with a 1024 channel scanner can guide microbubble-based therapy in volumetric regions of interest.

Keywords: CPS imaging; Volumetric ultrasound imaging; array design; image-guided therapy; microbubble imaging.

Figure 1

Field II simulations of 2.5 MHz pulse-echo (2-way) point spread functions (PSFs) at 10 cm depth for three array configurations with constant aperture size. Simulations of the azimuthal array: B-mode images for (A) 0.5λ - pitch array with 128 elements, (B) λ pitch array with 64 elements, and (C) 2λ - pitch with 32 elements; (D) Lateral resolution (cross-sections) of the PSFs. Simulation of the elevational array: B-mode images for (E) 3λ - pitch array with 8 elements, (F) 1.5λ - pitch array with 16 elements, and (G) 0.75λ - pitch with 32 elements, (H) Lateral resolution of the PSFs.

Figure 2

Field II simulations of 1.3 MHz transmitted (1-way) point spread functions (PSFs) at 3 cm depth for three array configurations with constant aperture size. Simulations of the azimuthal array: PSF intensities for (A) 0.31mm pitch array with 128 elements, (B) 0.62mm pitch array with 64 elements, and (C) 1.24mm pitch with 32 elements, (D) Lateral resolution of the PSFs. Simulation of the elevational array: PSF intensities for (E) 1.85 mm pitch array with 8 elements, (F) 0.92 mm pitch array with 16 elements, and (G) 0.46 mm pitch with 32 elements; (H) Lateral resolution of the PSFs.

Figure 3

Representation of the 3D volume extension resulting from plane wave imaging with 64 × 16 -element array: (A) photograph of the array and axis definition: x corresponding to the largest dimension of the array (azimuth), y, to the shortest, and z, to the propagation depth, (B) 1024-element array geometry, (C) geometrical definition of the plane wave angle, (D) azimuthal extension at imaging depth 10 cm, where α defines the azimuthal steering angle, (E) elevational extension at imaging depth 10 cm, where β defines the elevational steering angle. The 1024 channel, 4x256 imaging system configuration is shown in Fig 14.

Figure 4

Acoustic parameters of the 64 × 16 array pulse-echo measurement on a flat crystal: (A) center frequency map, (B) -6 dB bandwidth, (C) normalized sensitivity, (D) one-way time and frequency impulse responses were measured by placing a hydrophone at the natural focus and maximizing the received signal. Locally aberrant elements are dead elements that were not accounted for in the mean and standard deviation computations. a.u. = arbitrary units.

Figure 5

Two-way point spread function (PSF) estimation in the azimuthal dimension: (A) experimental PSFs obtained in the central azimuthal plane using a tungsten wire, (B) Verasonics (VSX) simulated PSFs using the central linear array (single column of the 2D transducer). Three different depths were tested in (A) and (B): 35 mm, 75 mm, and 110 mm. (C) Comparison of experimental lateral resolution of the PSFs. (D) Comparison of simulated lateral resolution of the PSFs.

Figure 6

Two-way point spread function (PSF) estimations in the elevational dimension: (A) experimental PSFs obtained in the central elevational plane using a 25-µm tungsten wire, (B) PSFs obtained in the central elevational plane with simulated 2D array. Three different depths were tested in (A) and (B): 35 mm, 75 mm, and 110 mm. Comparison of lateral resolution of the PSFs: (C) experiment; (D) simulation.

Figure 7

Effect of azimuthal and elevational steering on spherical cyst contrast (CIRS 050 3D phantom). Plane wave B-mode images in x, z plane with (A) azimuthal steering (AZ), (B) elevational steering (EL), (C) azimuthal and elevational steering compounded (AZ+EL). (D) Enlarged display of orange cysts (center line in azimuth). (E) Enlarged display of yellow cysts (off center). (F) Contrast ratio computation in two different phantom regions.

Figure 8

Three-dimensional B-mode image of a commercial phantom with inclusions of different echogenicity: Egg and cylinder phantom: (A) XZ central slice, (B) YZ central slice, (C) XY slice, (D) 3D rendering with reversed grayscale, (E) contrast comparison at two depths. Reflecting wires in speckle: (F) XZ central slice, (G) YZ central slice, (H) XY slice, (I) 3D rendering, (J) lateral resolution comparison at two depths.

Figure 9

Three-dimensional B-mode liver image of a healthy volunteer: (A) XZ central slice, (B) YZ central slice, (C) XY slice, (D) 3D rendering of the slice positions. (E) Table of volume and imaging rates for in vivo imaging of a 13×12×6cm³ volume, with a λ³ voxel size.

Figure 10

3D contrast imaging on a phantom setup. (A) Sketch of the experimental setup: silicone tubes for microbubble (MB) folw imaging are inserted orthogonally to each other in a polyvinyl alcohol (PVA) sponge. Slices of the CPS volume: (B) azimuthal XZ, (C) elevational YZ, (D) coronal XY. (E) 3D rendering of the CPS volume. Green, blue and red colored frames indicate the position of the slices respectively displayed in (B), (C), (D).

Figure 11

Therapeutic sequence. (A) Experimental timeline including CPS monitoring. (B) CPS images at different time points of exposure. (C) Higher intensity pixel count in the central azimuthal plane over time.

Figure 12

Microbubble monitoring. (A) B-mode image of silicone tubes in water with microbubbles flowing. (B) CPS image of silicone tubes in water with microbubbles flowing. (C) Fourier spectra of CPS signal with and without microbubble flow.

Figure 13

Ex vivo microbubble imaging in a mini-pig bone fracture model. (A) CPS image of central azimuth (x,z) plane (B) CPS image of central elevation (y,z) plane (C) Evolution of microbubble distribution through higher intensity voxel count over time - comparison of azimuth and elevation plane.

Figure 14

Global 1024-channel system architecture. Four 256 channel systems are combined. Each is composed of an ultrasound scanner and a host computer and equipped with a UTA 256 adaptor corresponding to one fourth of the array connectors. Secondary systems are defined and synchronized to an elected primary system.

Figure 15

Therapeutic sequence description. (A) Transmitted signal scheme displaying interleaving delays. (B) Final focal spot size representation after successive steered focused beams.