Martin Christian Hemmsen

An object-oriented multi-threaded software beamformation toolbox

Focusing and apodization are an essential part of signal processing in ultrasound imaging. Although the fundamental principles are simple, the dramatic increase in computational power of CPUs, GPUs, and FPGAs motivates the development of software based beamformers, which further improves image quality (and the accuracy of velocity estimation). For developing new imaging methods, it is important to establish proof-of-concept before using resources on real-time implementations. With this in mind, an effective and versatile Matlab toolbox written in C++ has been developed to assist in developing new beam formation strategies. It is a general 3D implementation capable of handling a multitude of focusing methods, interpolation schemes, and parametric and dynamic apodization. Despite being flexible, it is capable of exploiting parallelization on a single computer, on a cluster, or on both. On a single computer, it mimics the parallization in a scanner containing multiple beam formers. The focusing is determined using the positions of the transducer elements, presence of virtual sources, and the focus points. For interpolation, a number of interpolation schemes can be chosen, e.g. linear, polynomial, or cubic splines. Apodization can be specified by a number of window functions of fixed size applied on the individual elements as a function of distance to a reference point, or it can be dynamic with an expanding or contracting aperture to obtain a constant F-number, or both. On a standard PC with an Intel Quad-Core Xeon E5520 processor running at 2.26 GHz, the toolbox can beamform 300.000 points using 700.000 data samples in 3 seconds using a transducer with 192 elements, dynamic apodization in transmit and receive, and cubic splines for interpolation. This is 19 times faster than our previous toolbox.

Implementation of a versatile research data acquisition system using a commercially available medical ultrasound scanner

This paper describes the design and implementation of a versatile, open-architecture research data acquisition system using a commercially available medical ultrasound scanner. The open architecture will allow researchers and clinicians to rapidly develop applications and move them relatively easy to the clinic. The system consists of a standard PC equipped with a camera link and an ultrasound scanner equipped with a research interface. The ultrasound scanner is an easy-to-use imaging device that is capable of generating high-quality images. In addition to supporting the acquisition of multiple data types, such as B-mode, M-mode, pulsed Doppler, and color flow imaging, the machine provides users with full control over imaging parameters such as transmit level, excitation waveform, beam angle, and focal depth. Beamformed RF data can be acquired from regions of interest throughout the image plane and stored to a file with a simple button press. For clinical trials and investigational purposes, when an identical image plane is desired for both an experimental and a reference data set, interleaved data can be captured. This form of data acquisition allows switching between multiple setups while maintaining identical transducer, scanner, region of interest, and recording time. Data acquisition is controlled through a graphical user interface running on the PC. This program implements an interface for third-party software to interact with the application. A software development toolkit is developed to give researchers and clinicians the ability to utilize third-party software for data analysis and flexible manipulation of control parameters. Because of the advantages of speed of acquisition and clinical benefit, research projects have successfully used the system to test and implement their customized solutions for different applications. Three examples of system use are presented in this paper: evaluation of synthetic aperture sequential beamformation, transverse oscillation for blood velocity estimation, and acquisition of spectral velocity data for evaluating aortic aneurysms.

Ultrasound image quality assessment: a framework for evaluation of clinical image quality

Improvement of ultrasound images should be guided by their diagnostic value. Evaluation of clinical image quality is generally performed subjectively, because objective criteria have not yet been fully developed and accepted for the evaluation of clinical image quality. Based on recommendation 500 from the International Telecommunication Union - Radiocommunication (ITU-R) for such subjective quality assessment, this work presents equipment and a methodology for clinical image quality evaluation for guiding the development of new and improved imaging. The system is based on a BK-Medical 2202 ProFocus scanner equipped with a UA2227 research interface, connected to a PC through X64-CL Express camera link. Data acquisition features subject data recording, loading/saving of exact scanner settings (for later experiment reproducibility), free access to all system parameters for beamformation and is applicable for clinical use. The free access to all system parameters enables the ability to capture standardized images as found in the clinic and experimental data from new processing or beamformation methods. The length of the data sequences is only restricted by the memory of the external PC. Data may be captured interleaved, switching between multiple setups, to maintain identical transducer, scanner, region of interest and recording time on both the experimental- and standardized images. Data storage is approximately 15.1 seconds pr. 3 sec sequence including complete scanner settings and patient information, which is fast enough to get sufficient number of scans under realistic operating conditions, so that statistical evaluation is valid and reliable.

Implementation of synthetic aperture imaging on a hand-held device

This paper presents several implementations of Synthetic Aperture Sequential Beamforming (SASB) on commercially available hand-held devices. The implementations include real-time wireless reception of ultrasound radio frequency signals and GPU processing for B-mode imaging. The proposed implementation demonstrates that SASB can be executed in-time for real-time ultrasound imaging. The wireless communication between probe and processing device satisfies the required bandwidth for real-time data transfer with current 802.11ac technology. The implementation is evaluated using four different hand-held devices all with different chipsets and a BK Medical UltraView 800 ultrasound scanner emulating a wireless probe. The wireless transmission is benchmarked using an imaging setup consisting of 269 scan lines × 1472 complex samples (1.58 MB pr. frame, 16 frames per second). The measured data throughput reached an average of 28.8 MB/s using a LG G2 mobile device, which is more than the required data throughput of 25.3 MB/s. Benchmarking the processing performance for B-mode imaging showed a total processing time of 18.9 ms (53 frames/s), which is less than the acquisition time (62.5 ms).

System-Level Design of an Integrated Receiver Front End for a Wireless Ultrasound Probe

In this paper a system-level design is presented for an integrated receive circuit for a wireless ultrasound probe, which includes analog front-ends and beamformation modules. The study focuses on the investigation of the effects of architectural design choices on the image quality. The point spread function is simulated in Field II from 10 to 160mm using a convex array transducer. A noise analysis is performed, and the minimum signal-to-noise ratio (SNR) requirements are derived for the low-noise amplifiers (LNAs) and A/D converters (ADCs) to fulfil the design specifications of a dynamic range of 60 dB and a penetration depth of 160mm in the B-mode image. Six front-end implementations are compared using Nyquist-rate and modulator ADCs. The image quality is evaluated as a function of the depth in terms of lateral full-width at half maximum (FWHM) and -12 dB cystic resolution (CR). The designs that minimally satisfy the specifications are based on a 8-bit 30MSPS Nyquist converter and a single-bit 3rd order 240 MSPS modulator, with a SNR for the LNA in both cases equal to 64 dB. The mean lateral FWHM and CR are 2.4% and 7.1% lower for the architecture compared to the Nyquistrate one. However, the results generally show minimal differences between equivalent architectures. Advantages and drawbacks are finally discussed for the two families of converters.

Tissue harmonic synthetic aperture ultrasound imaging

Synthetic aperture sequential beamforming (SASB) and tissue harmonic imaging (THI) are combined to improve the image quality of medical ultrasound imaging. The technique is evaluated in a comparative study against dynamic receive focusing (DRF). The objective is to investigate if SASB combined with THI improves the image quality compared to DRF-THI. The major benefit of SASB is a reduced bandwidth between the probe and processing unit. A BK Medical 2202 Ultraview ultrasound scanner was used to acquire beamformed RF data for offline evaluation. The acquisition was made interleaved between methods, and data were recorded with and without pulse inversion for tissue harmonic imaging. Data were acquired using a Sound Technology 192 element convex array transducer from both a wire phantom and a tissue mimicking phantom to investigate spatial resolution and penetration. In vivo scans were also performed for a visual comparison. The spatial resolution for SASB-THI is on average 19% better than DRI-THI, and the investigation of penetration showed equally good signal-to-noise ratio. In vivo B-mode scans were made and compared. The comparison showed that SASB-THI reduces the artifact and noise interference and improves image contrast and spatial resolution.

Implementation of real-time duplex synthetic aperture ultrasonography

This paper presents a real-time duplex synthetic aperture imaging system, implemented on a commercially available tablet. This includes real-time wireless reception of ultrasound signals and GPU processing for B-mode and Color Flow Imaging (CFM). The objective of the work is to investigate the implementation complexity and processing demands. The image processing is performed using the principle of Synthetic Aperture Sequential Beamforming (SASB) and the flow estimator is implemented using the cross-correlation estimator. Results are evaluated using a HTC Nexus 9 tablet and a BK Medical BK3000 ultrasound scanner emulating a wireless probe. The duplex imaging setup consists of interleaved B-mode and CFM frames. The required data throughput for real-time imaging is 36.1 MB/s. The measured data throughput peaked at 39.562 MB/s, covering the requirement for real-time data transfer and overhead in the TCP/IP protocol. Benchmarking of real-time imaging showed a total processing time of 25.7 ms (39 frames/s) which is less than the acquisition time (29.4 ms). In conclusion, the proposed implementation demonstrates that both B-mode and CFM can be executed in-time for real-time ultrasound imaging and that the required bandwidth between the probe and processing unit is within the current Wi-Fi standards.

Novel automatic detection of pleura and B-lines (comet-tail artifacts) on in vivo lung ultrasound scans

This paper presents a novel automatic method for detection of B-lines (comet-tail artifacts) in lung ultrasound scans. B-lines are the most commonly used artifacts for analyzing the pulmonary edema. They appear as laser-like vertical beams, which arise from the pleural line and spread down without fading to the edge of the screen. An increase in their number is associated with presence of edema. All the scans used in this study were acquired using a BK3000 ultrasound scanner (BK Ultrasound, Denmark) driving a 192-element 5:5 MHz wide linear transducer (10L2W, BK Ultrasound). The dynamic received focus technique was employed to generate the sequences. Six subjects, among those three patients after major surgery and three normal subjects, were scanned once and Six ultrasound sequences each containing 50 frames were acquired. The proposed algorithm was applied to all 300 in-vivo lung ultrasound images. The pleural line is first segmented on each image and then the B-line artifacts spreading down from the pleural line are detected and overlayed on the image. The resulting 300 images showed that the mean lateral distance between B-lines detected on images acquired from patients decreased by 20% in compare with that of normal subjects. Therefore, the method can be used as the basis of a method of automatically and qualitatively characterizing the distribution of B-lines.

Synthetic Aperture Sequential Beamforming implemented on multi-core platforms

This paper compares several computational approaches to Synthetic Aperture Sequential Beamforming (SASB) targeting consumer level parallel processors such as multi-core CPUs and GPUs. The proposed implementations demonstrate that ultrasound imaging using SASB can be executed in real-time with a significant headroom for post-processing. The CPU implementations are optimized using Single Instruction Multiple Data (SIMD) instruction extensions and multithreading, and the GPU computations are performed using the APIs, OpenCL and OpenGL. The implementations include refocusing (dynamic focusing) of a set of fixed focused scan lines received from a BK Medical UltraView 800 scanner and subsequent image processing for B-mode imaging and rendering to screen. The benchmarking is performed using a clinically evaluated imaging setup consisting of 269 scan lines × 1472 complex samples (1.58 MB per frame, 16 frames per second) on an Intel Core i7 2600 CPU with an AMD HD7850 and a NVIDIA GTX680 GPU. The fastest CPU and GPU implementations use 14% and 1.3% of the real-time budget of 62 ms/frame, respectively. The maximum achieved processing rate is 1265 frames/s.

Preliminary study of synthetic aperture tissue harmonic imaging on in-vivo data

A method for synthetic aperture tissue harmonic imaging is investigated. It combines synthetic aperture sequen- tial beamforming (SASB) with tissue harmonic imaging (THI) to produce an increased and more uniform spatial resolution and improved side lobe reduction compared to conventional B-mode imaging. Synthetic aperture sequential beamforming tissue harmonic imaging (SASB-THI) was implemented on a commercially available BK 2202 Pro Focus UltraView ultrasound system and compared to dynamic receive focused tissue harmonic imag- ing (DRF-THI) in clinical scans. The scan sequence that was implemented on the UltraView system acquires both SASB-THI and DRF-THI simultaneously. Twenty-four simultaneously acquired video sequences of in-vivo abdominal SASB-THI and DRF-THI scans on 3 volunteers of 4 different sections of liver and kidney tissues were created. Videos of the in-vivo scans were presented in double blinded studies to two radiologists for image quality performance scoring. Limitations to the systems transmit stage prevented user defined transmit apodization to be applied. Field II simulations showed that side lobes in SASB could be improved by using Hanning transmit apodization. Results from the image quality study show, that in the current configuration on the UltraView system, where no transmit apodization was applied, SASB-THI and DRF-THI produced equally good images. It is expected that given the use of transmit apodization, SASB-THI could be further improved.