Search

3D Ultrasound Computed Tomography (USCT) is an imaging method for early breast cancer detection. The third generation 3 USCT device is developed at Karlsruhe Institute of Technology. The USCT III device has a hemispherical transducer distribution and emits and receives nearly spherical waves. This enables reflection and transmission imaging simultaneously and fully in 3D. The main challenges for the front-end design are to integrate a large number of transducers, to allow high voltage coded excitation, and to receive low amplitude signals with high quality. These challenges were solved using a smart sensor frontend design with a custom application specific integrated circuit (ASIC).

Conventional iterative methods for image reconstruction in ultrasound transmission tomography need to perform many iterations where at each iteration one has to compute the complex forward model of ultrasound wave propagation, and hence they are time-consuming. We use a U-net neural network to accelerate the reconstruction, by training the network to map from an initial reconstruction obtained via a few iterations of L-BFGS method to the target ground truth image. Since the computation of a forward pass of the neural network is very fast, we can expect a significant acceleration using the trained network for image reconstruction. Experiments show that our trained network can replace 40 L-BFGS iterations to generate equivalent reconstructions with slightly better quantitative quality in terms of normalized root mean square error and better visual quality due to the network's denoising effect. It can achieve up to 283× speedup compared with L-BFGS method for reconstructing small-size sound speed images with 80×80 pixels. This implies that we can expect even greater acceleration effects when applying such approach to reconstruct large-size 3D images.

A semi-automated packaging process for transducer array manufacturing is presented. As an major innovation, a transducer disc is integrated in a sandwich structure between a printed circuit board (PCB) and an acoustic matching layer. Each of the transducer discs contains 18 lead-zirconium-titanate (PZT) fibres embedded in epoxy. To interconnect the transducer array components, adhesive bonding and automatic pick-and-place processes were used. A pre-series was evaluated by measuring the electro-mechanical impedance (EMI) before and after the assembly. Statistical analysis showed consistent behaviour of the series resonance f s and the electro-mechanical coupling k eff before and after the packaging. This encouraged the manufacturing of 256 arrays. These arrays will now be integrated in an ultrasound computer tomography (USCT) system with 3D scanning aperture for breast cancer imaging. With this system, we intend to bridge the gap towards clinical use of full 3D USCT.

The requirements of our third generation 3D ultrasound computer tomography system (USCT) led to a transducer array design based on single-fibre piezocomposites. The used piezoelectric ceramic fibres exhibit a thickness to diameter ratio of 4/3. This ratio results in a proximity of the dominant axial and lateral vibration resonances. The standard Krimholtz-Leedom-Matthaei (KLM) model considers only one spatial dimension. This limits the validity of the modelled predictions for our application. By adding a coupled transmission line in series to the axial resonator, the standard model can be extended to take additional spatial dimensions into account. Using model parameters from literature resulted in a low model-to-measurement fit. To enhance the prediction accuracy, a brute-force optimisation over four model parameters was conducted. With the optimal parameter set, a fit of 90.3% could finally be achieved. The necessary model parameter changes were found reasonable for the piezocomposite design. With the presented model enhancement it is now possible to consider both dominant vibration modes of our transducers.

An airborne ultrasound imaging system was developed for reflection tomography. The ultrasound transducers surround the region of interest (ROI) in an arrangement optimized for maximum coverage and homogeneous distributed image quality. In this work, we developed a workflow for automatic segmentation and classification of objects in the reconstructed images. Our workflow can be applied for varying intensities of object edges with a local maxima based segmentation and a multi-parameter image reconstruction. The segmented regions are classified with a neural network, and the object localization was implemented with Generalized Hough Transform using a custom template for each classified object in the data set. A classification accuracy of 95% for six trained test objects and a localization accuracy of 5 mm were achieved.

Airborne ultrasonic sensors are widely used in industry for detecting movement of targets and measuring their distances to the sensor. More complex object localization is usually done with optical systems, yet their application is limited. In this work an airborne ultrasound imaging system is developed, which reconstructs an image of the region-of-interest (ROI) with multiple objects. With further image processing, this system allows more sophisticated object localization. The proposed optimization method results in a sensor system with minimum number of sensors and improved sensor distribution for maximum coverage of the ROI and homogeneous performance.

For iterative image reconstruction of transmission tomography we apply the paraxial approximation of the Helmholtz equation for a spherical transducer arrangement. We choose this approach due to its three order of magnitude lower complexity than full wave solutions with the same precision for transmission tomography. In homogeneous media we prove that our forward solution is exact. With the help of this forward solution 2D and 3D ultrasound measurements could be simulated for transmission tomography. 2D reconstructions of a breast-like numerical phantom had a deviation in sound speed of 0.14 m/s and a deviation in attenuation of 6.5% from the ground truth. Applications up to now are breast cancer diagnostics and non-destructive testing.

A promising candidate for breast cancer imaging is ultrasound computer tomography (USCT). At Karlsruhe Institute of Technology (KIT) a 3D USCT with a semiellipsoidal aperture consisting of several hundreds of US transducers was built. The utilized imaging methods require accurate and reliable temperature information over the measured space and acquisition time. Several methods were applied which improved the temperature accuracy and reliability beyond the hardware defaults by one order of magnitude from ±1°C to approx. 0.005°C beyond the requirement of 0.1°C.

A device for 3D ultrasound computer tomography system is currently under development at KIT with the goal of high-resolution images for early breast cancer detection. With its semi-ellipsoidal positioning of 2041 ultrasound transducers around the breast in space, full 3D images can be reconstructed. Calibration process of such a complex system is very time-consuming and difficult. This paper proposes a Newton's method based self calibration using time-of-flight measurements between each emitter and receiver. One unique feature of this method is the separation of each potential error sources in the system and sequential calibration according to their magnitudes. This enables the analysis of each error component in the system. Simulation and application on real data have both shown sub-wavelength accuracy in the calibration results.

3D Ultrasound Computer Tomography (USCT) aims at early detection of breast cancer. The system was optimized for Synthetic Aperture Focusing Technique (SAFT) using unfocused ultrasound transducers. Due to low SNR of the signals, Coded Excitation (CE) and Matched Filter (MF) are applied. For complex objects the received signals (A-scans) consist of multiple overlaid reflections. The overlaid reflections are difficult to separate due to the system's limited bandwidth which results in side lobes. Non-separable reflections limit the resolution and contrast of the resulting images. Using Huffman Sequences (HS) as CE, a trade-off can be set between the side lobe suppression and loss of main lobe energy. Higher suppression of side lobes results in reduced energy in the CE, i.e. smaller SNR. So an approach has been implemented which suppresses the side lobes, but preserves the energy of CE by designing a Huffman sequences based mismatched filter. Combination of a CE and different filters were evaluated for single reflection and multiple overlaid reflections. The results show high side lobe suppression within limited bandwidth and preserve the SNR.

The 3D USCT II system needs a method to localize and reposition a non-centrally positioned breast. A novel imaging technique (IIT) was introduced to retrieve intersection images of breasts. By detecting the breast deviation direction from these images, the breast position can be estimated. This method limits the amount of data and speeds up the data transmission by producing only 2D images and generalizing the breast as convex. It achieves an mean accuracy of 0.73° for the estimation of the breast deviation direction. IIT achieves a data reduction rate of 99.99%, which speeds up the data transfer to 0.1 s. Therefore, IIT achieves a fast and sufficiently accurate detection of the breast position for USCT II system.

In a first pilot study with 3D Ultrasound Computer Tomography especially the speed of sound images of the breast showed promising results for breast cancer detection. Yet, the resolution of transmission tomography in our system is limited in comparison to the reflectivity volumes. In this paper we describe a setup for an improved device overcoming this limitation. An optimized geometry of the transducers in a sparse distribution improves both the contrast in reflection and contrast and resolution in transmission tomography by nearly an order of magnitude. Furthermore data acquisition is accelerated from 10 to less than 2 minutes. The derived methods, simulation, and measuring results are described.

The KIT 3D USCT was tested in a pilot study on ten patients. The primary goals of the pilot study were to test the USCT device, the data acquisition protocols, the image reconstruction methods and the image fusion techniques in a clinical environment. The study was conducted successfully; the data acquisition could be carried out for all patients with an average imaging time of six minutes per breast. First reconstructions provide promising images. Overlaid volumes of the modalities show qualitative and quantitative information at a glance. The results led to further optimization of the system and the data acquisition protocol.

A promising candidate for breast cancer imaging is ultrasound computer tomography (USCT). At KIT a 3D USCT with a semi-ellipsoidal aperture consisting of several hundreds of US transducers was built. Spherical waves are sequentially emitted and received. 3D SAFT is applied to reconstruct reflectivity volumes. However, straight forward SAFT imaging leads to blurred images. A post-imaging de-blurring approach applying aperture characteristics is described here and analyzed with a simulation and a clinical data set. A increase of 26% in the mean resolution in imaging simulation with a point scatters was achieved. A increase in resolution with breast images was observed, too.

3D ultrasound computer tomography (3D USCT) promises reproducible high-resolution images for early detection of breast tumors. The KIT 3D USCT provides three different modalities (reflectivity, speed of sound, and attenuation) using 2041 transducers. In this setup, with a diameter of 26 cm and height of 17 cm, ultrasound can travel over long distances up to 52 cm. Phase aberrations (PA) due to speed of sound (SOS) variations inside the measuring object (water, different breast tissues) cause many pulses not to overlap in a distinct voxel for the coherent reflectivity reconstruction. Previous research showed that image quality can be increased significantly performing a PA correction. As no quantitative error assessment was done yet, a simulation based on ray tracing is used to quantify their image degradation caused by PA and the effects of the applied PA correction. This was done with the metrics contrast, resolution and displacement for different positions in the 3D USCT. Our work shows that PA correction significantly restores the image quality.

Ultrasound Computer Tomography (USCT) is a promising 3D modality for early breast cancer detection, which is expected to provide quantitative imaging. The aim of this paper is to evaluate the quantitative diagnostic value of the USCT images, i.e. sound speed and attenuation images, using X-ray mammograms as ground truth. For this purpose we applied our 2D/3D registration method, which is based on biomechanical modeling of the breast. Mammograms were segmented into fatty, glandular and tumorous tissue. For each tissue, the average sound speed and attenuation in the corresponding USCT images was calculated. Tumorous tissue could be separated from fatty and glandular tissue using a fixed absolute sound speed threshold in all regarded datasets. By combining sound speed and attenuation, the separation between fatty and glandular tissue could be improved. By overlaying sound speed and attenuation information on the mammogram, quantitative and morphological information can be combined for multimodal diagnosis. This may benefit early breast cancer detection in future.

With the rise of heterogeneous computing architectures, application developers are confronted with a multitude of hardware platforms and the challenge of identifying the most suitable processing platform for their application. Strong competitors for the acceleration of 3D Ultrasound Computer Tomography, a medical imaging method for early breast cancer diagnosis, are GPU and FPGA devices. In this work, we evaluate processing performance and efficiency metrics for current FPGA and GPU devices. We compare top-notch devices from the 40 nm generation as well as FPGA and GPU devices, which draw the same amount of power. For our two benchmark algorithms, the results show that if power consumption is not considered the GPU and the FPGA give both, a similar processing performance and processing efficiency per transistor. However, if the power budget is limited to a similar value, the FPGA performs between six and eight times better than the GPU.

Calculating and visualizing the sound field of transducers is a key task in the development of most ultrasound applications. In most cases it is sufficient to calculate the directivity pattern or a plane slice of the sound field using the point source synthesis method. To date no freeware tool is available which is able to simulate the sound field for arbitrary rectangular transducer array matrices for different excitation signals. The developed ”Transducer Array Calculation GUI” (TAC) includes phase- and frequency-informations as well as apodization to calculate sound fields of various transducer configurations. TAC also offers to include the medium attenuation and transducer's electrical impedance characteristics. Our software supports fast and realistic transducer engineering and it is open source.

We designed and built a 3D ultrasound computer tomography (USCT) device with a nearly isotropic and spatially invariant 3D point spread function, to be tested in a clinical study. The objective of this work was to image two healthy volunteers and to evaluate the USCT volumes in comparison to corresponding Magnetic Resonance Images (MRI). The here presented volumes are reflectivity images generated with 3D synthetic aperture focusing technique. The volunteers were imaged with different parameterizations of the data acquisition. The data acquisition time was between four and twelve minutes. For both volunteers we found that the breast surface and inner structures are clearly shown in the USCT volume and fit the structures given by the MRI.

We have developed a three-dimensional ultrasound computer tomography system (3D USCT) for early breast cancer diagnosis. The method gives reproducible volume images of the female breast in 3D. The optimized aperture and transducer array systems show at the used low ultrasound frequency of 2.5 MHz a very high image quality. In contrast to other methods our system has a nearly isotropic and spatially invariant 3D point spread function (PSF) with a resolution of 240 11m as measured with a 70 11m AI bond wire. By simulated spectra including the experimentally measured noise of transmission tomography it could be shown that structures down to 4 mm could be reconstructed in speed of sound images. The reconstruction of the 3D images in reflection tomography could be accelerated by the use of the multi-core structure of PCs and eight graphical processor units by a factor of 94. For a reconstruction of the reflectivity with 64 layers of 10242 pixels we reduce the computing time from 33 hours to about 21 minutes. In a first clinical study and the images of two healthy probands the observed structures in breasts due to 3D USCT are comparable to images obtained with a clinical MRT.

Ultrasound Computer Tomography (USCT) is an upcoming modality for early breast cancer diagnosis, which provides morphological as well as quantitative imaging. In order to compare USCT images to the standard modality X-ray mammography, a 2D/3D registration has to be applied. To analyze the relevance of sound speed images as a quantitative imaging method for tissue characterization, the aim of this paper is to quantify sound speed values of different types of tissue using mammograms as ground truth. Mammograms are segmented and classified into fat, glandular and tumorous tissue. For each tissue class the average sound speed in the registered sound speed image is calculated. The mean absolute sound speed was 1457 m/s for regions segmented as fatty tissue, 1470 m/s for glandular and 1509 m/s for tumorous tissue. For all ten datasets, the sound speed in tumorous tissue was significantly higher than in glandular and fatty tissue. Also glandular and fatty tissue could be separated easily by the absolute sound speed values. By color-coding sound speed, quantitative information from USCT and morphological information from X-ray mammography are fused for combined diagnosis. We believe this method will help radiologists in gaining experience in the reading of USCT images. The combination of diagnostic information is likely to be beneficial to early breast cancer detection.

3D ultrasound computer tomography (USCT) is a new and promising method for early breast cancer diagnosis. An ultrasound computer tomograph was developed by the Karlsruhe Institute of Technology (KIT) and provides a resolution of 0.2 mm. The main components are the semi-ellipsoidal aperture with 628 ultrasound (US) emitters and 1413 US receivers and the 480 input channel data acquisition system. An additional external computing system is used for the time consuming image reconstruction. To reduce the reconstruction time different implementations on massive parallel computing architectures like multicore processor, GPUs and FPGAs were tested. One realization applies reconfiguration of the 60 FPGAs inside the DAQ system. The investigations show significant acceleration of the reconstruction time up to a factor of 15.8 for the latest FPGA generation and 17.6 for a state of the art GPU.

As today's standard screening methods frequently fail to diagnose breast cancer before metastases have developed, earlier breast cancer diagnosis is still a major challenge. Three-dimensional ultrasound computer tomography promises high-quality images of the breast, but is currently limited by a time-consuming synthetic aperture focusing technique based image reconstruction. In this work, we investigate the acceleration of the image reconstruction by a GPU, and by the FPGAs embedded in our custom data acquisition system. We compare the obtained performance results with a recent multi-core CPU and show that both platforms are able to accelerate processing. The GPU reaches the highest performance. Furthermore, we draw conclusions in terms of applicability of the accelerated reconstructions in future clinical application and highlight general principles for speed-up on GPUs and FPGAs.

Our 3D Ultrasound Computer Tomograph for breast imaging uses unfocused transducers grouped on a large non-planar aperture and synthetic aperture focusing in 3D. Technical feasibility limits the number of transducer positions to a sparse aperture, causing artifacts due to grating lobe effects in the resulting images. To suppress the artifacts, a median filter is applied to special pressure distributions, which are derived for each image point. After filtering the distributions are summed up for reconstruction. This approach reduces the artifacts significantly. For a phantom of ten nylon threads (0.2 mm diameter) an increase of contrast (SDNR) from 7.6 to 35.9 could be achieved. Also more complex phantoms showed significant reduction. The presented approach is simple and very effective, and can be easily integrated into the applied synthetic aperture focusing technique.

We describe the second generation of a 3D-Ultrasound Computer Tomography (USCT) system. After we achieved in the first generation a device with sub-wavelength resolution and three imaging modalities (reflection, attenuation, speed of sound) and tested it with static phantoms, we developed a device for in-vivo imaging. In the new system the geometry of transducers and their spatial distribution is optimized in respect to uniformity and high value of: contrast, resolution, and illumination. Furthermore we developed new electronics which allows faster DAQ (≤ 2 min) and contains larger and faster FPGAs to use their processing power for data pre-processing.

3D ultrasound computer tomography (3D USCT) is a new imaging method aimed at early breast cancer detection. For synthetic aperture focusing the data is acquired with (nearly) unfocussed ultrasound emission and reception. Therefore the SNR of the data is low and needs to be optimized.

Ultrasound Computer Tomography (USCT) records the interaction of ultrasonic waves with an imaged object from many different angles. The main advantage of such a system is simultaneous recording of reflection, absorption and speed of sound images, and higher image quality with fast data acquisition. Until now, it was not feasible to build a 3D USCT due to the high demands on the required number of transducers, high data rate, and time consuming post-processing. The aim of this first experimental setup for 3D USCT was to analyze the feasibility for breast cancer diagnosis of such a system built with today’s technology and draw conclusions for future setups. The results of this system are presented and discussed in regard of data acquisition, image quality, and duration of image reconstruction. The main conclusion is that 3D USCT is feasible with today’s technology, if a sparse aperture or long DAQ times can be accepted. A setup for a second-generation 3D USCT with the main aim to image volunteers and carry out a small preclinical study is proposed.

At Forschungszentrum Karlsruhe, a new imaging system for early diagnosis of breast cancer is currently developed. The 3D Ultrasound Computer Tomography (USCT) is based on approx. 2000 ultrasound transducers which produce 20 GB of raw data consisting of 3.5 mio. A-scans (amplitude scans) for one image. The large number of A-scans, the large amount of data and the complex relationship between raw data and reconstructed image makes analysis, understanding and further development difficult for the scientists and especially for new employees and students. For this reason, an interactive graphical user interface (GUI) was developed using MATLAB. It integrates existent analysis methods and is easily extendable with new functionality via a plugin concept. The software provides several visualization functions for the raw data, the reconstructed 3D images, the USCT aperture and the relationships between them. The software shows the capability of MATLAB as programming language for numerical solutions as well as GUIs and complex software systems. It has a large benefit to the working group by integrating analysis methods and sharing them. The GUI and the visualization of the complex relationships of the USCT reduces the training period for new employees and students.

At Forschungszentrum Karlsruhe a 3D ultrasound computer tomograph (USCT) for breast cancer diagnosis is currently under development. For many applications, i.e. reconstruction of speed of sound maps, signal denoising,signal compression and calibration, it is necessary to detect pulses and their parameters (center frequency, bandwidth factor, time of arrival, phase and amplitude) accurately and robustly. The pulse detection is demanding due to low signal-to-noise ratios (SNR) caused by unfocused pulses from single emitters. Besides that angle dependent pulse shapes result in ultrasonic echoes which are similar in their center frequency but vary strongly in their bandwidth factor. Additionally, in sum 3.5 million A-scans (20 GB) are acquired, so that the pulse detection has to be fast and efficient. This is achieved by using a fast pre-classifler in order to separate the ultrasonic echoes from the non-white system noise of our system. The detected echoes are then passed individually to a parameter estimation method to determine pulse parameters accurately. An analysis of several classifiers resulted in an alternating decision tree which is both fast and accurate. Classification performance of 95% could be achieved as well as robust parameter estimation with non-white system noise if the SNR is larger than 3 dB. A comparative image reconstruction resulted in significantly sharper images.

Our experimental setup for ultrasound computer tomography (USCT) has a cylindrical aperture with a diameter of 18 cm and height of 15 cm. In this setup ultrasound can travel over long distances up to 50 cm. Phase aberration due to speed of sound variations cause many pulses not to overlap in the reflectivity reconstruction. 3D speed of sound (SOS) maps were calculated using the recorded data and applied for correction. Due to a sparse sampling of the received transmission signals, the resulting speed of sound maps are noisy and of low detail. But the accuracy of the average speed of sound is sufficient for phase aberration correction of the reflection images down to object size of approx. 1 mm and increases the image quality significantly. Our work shows that high resolution USCT is not possible without phase aberration correction using SOS information.

For our current 3D demonstrator system for ultrasound computer tomography (USCT) a synthetic aperture based image reconstruction is implemented whose calculation time is in the order of days. Faster approximations might lead to large errors for our USCT setup which uses many multistatic emitter-receiver combinations. The aim of this work is the analysis of non-linear signal to image mapping for a cylindrical aperture and quantify the errors of computational less complex approximations. Four approaches with increasing computational complexity were implemented, tested with the cylindrical setup and compared to a commonly used linear transducer setup. For the linear aperture the second approximation leads to satisfying results. For the more diverging USCT setup a more accurate approximation is needed: the third approximation, which is still less computational expensive as the ground-truth implementation, gives satisfying results.

The contribution presents some results obtained on the way to checking (and partly complementing) the standard reconstruction procedures in USCT by wave-equation based simulations. Mathematical models emerging from the transparent physical background for both the surrounding fluid and the object tissue are presented, followed by the present results of developing a realistic original finite-element- method based simulation. With respect to the need of comparison with the calibration measurements, a preliminary optimization of initial guesses to boundary conditions at the transducer array is discussed, based on a point-source model. The computational requirements of the procedures are also mentioned together with concrete examples of achieved results.

Ultrasound computer tomography (USCT) is an ultrasound imaging method capable of producing volume images with high spatial resolution and image quality. The long term goal of the system under construction is high quality imaging in 3D for early breast cancer diagnosis. In this paper the first images of an experimental 3D-prototype are presented and discussed in respect to further modifications of our system.

Ultrasound computer tomography (USCT) is an imaging method capable of producing volume images with sub-millimeter resolution and high image quality. The long term goal of the system under construction is 3D imaging for early breast cancer diagnosis. In this paper the first images of a clinical breast phantom are presented and discussed in respect to further modifications of our system

At Forschungszentrum Karlsruhe a 3D ultrasound computer tomograph (USCT) is currently under construction aimed at early breast cancer diagnosis. Ellipsoidal backprojection is applied to reconstruct 3D images. As USCT is a system used for imaging, there is a profound interest in analyzing the image resolution and image quality of this system, especially with respect to its system parameters such as number of the sensors, the geometry of the aperture, applied signal processing, etc. In this paper the results of first assessments of the resolution are presented. The resolution is evaluated in terms of the 2D full width at half maximum (2D FWHM) of the system's 3D point spread function (PSF). Slice images of 3D PSFs have been calculated to assess the spatial variability of the PSF within the current setup of the system. However, calculating these results is very time-consuming, thus an approximative approach for the analysis of the 3D PSF in XY and Z has been developed. This approach allows the analysis of the PSF while changing a variety of system parameters in real time. The long term goal of this project is the optimization of the system parameters with respect to the image quality