William F. Walker, Associate Professor, and Francesco Viola, Research Associate, along with a graduate student Michael Ellis, at the University of Virginia Engineering School, have developed an advanced imaging algorithm that improves resolution of images in ultrasound. The biomedical engineering team has come up with an innovative method of signal processing, which can be used in a wide range of imaging and sensing systems, apart from its use in medical ultrasounds.
The algorithm, named as Time-domain Optimized Near-field Estimator (TONE), is designed to improve the image resolution and contrast during medical imaging procedures. The computer algorithm in an ultrasound scanner makes use of reflected sound waves to create real-time images of the organ or tissue that is being imaged, but these images are not always clear. The off-axis signals (reflections coming from undesired locations) result in poor quality of images produced by the ultrasound systems. TONE reportedly reduces these unwanted signals, resulting in an image with high resolution and contrast.
The team performed a series of simulations using sample ultrasound data to test the performance of TONE and compared the results with conventional beamforming (CBF) strategies. They conducted imaging trials by suspending the wires in water, which is a typical set up to test image resolution and contrast in medical ultrasound instruments. The tests showed that TONE increased the spatial resolution during imaging compared to CBF. The team made use of an interactive parallel computing platform Star-P, developed by Interactive Supercomputing Corporation, to handle the computing processes during the experiments.
Although the algorithm has not yet been put to use for imaging human tissues, the next intended step of its use is just that. Researchers believe that TONE not only plays a role in improving the ultrasound images, but also can be applied in the field of radio astronomy, seismology, telecommunications, etc.
In the 12th Institute of Electrical and Electronics Engineers, Inc (IEEE) symposium on Computer-Based Medical Systems, 1999 Sapia et al presented their study on improving the quality of ultrasound image using adaptive inverse filtering. They reported that the resolution of clinical ultrasound image is reduced as a result of limited effective aperture size (convolutional blurring), out-of-focus blurring, and noise. 3D microscope images were deconvolved using an adaptive, least-mean-square solution to a statistical Wiener filter. This filter could also be applied to 2D and 3D images as a finite impulse response (FIR) filter using the inverse model of point-spread-function (PSF). For ultrasound images, the filter could be solved using the response to a phantom, with the desired result known before hand, and it was done adaptively to reduce the mean-square-error. The filter that was resolved in this manner could be applied to any image acquired with the same transducer array and instrument parameters as those used during solving process. The preliminary results of using such filters had shown promising results of improving resolution and minimizing the noise in ultrasound images.
Sun Kim and Beom Ra (2005) had earlier come up with an image enhancement algorithm based on a multi-resolution approach for 2D B-mode ultrasound images. In their algorithm, Kim and Ra performed directional filtering and noise reducing procedures from the coarse to fine resolution images that were obtained from wavelet-transformed data. For directional filtering, they used Eigen-analysis to examine the structural feature at each pixel, and if the pixel belonged to the edge region, they performed directional smoothing to improve its continuity, and directional sharpening to enhance the contrast of the image. Speckle noise was alleviated by reducing the wavelet coefficients corresponding to the homogeneous region, which was determined by considering the frequency characteristics of the speckle. Results obtained by their studies showed that the algorithm improved the quality of ultrasound image without generating any noticeable artifacts.
Improving the resolution and contrast of ultrasound images will help physicians in diagnosing illnesses with better accuracy. Although the development of software like TONE is a breakthrough in medical imaging technology, it needs better adaptation and fine-tuning before it can be put into use at a clinical level.
About University of Virginia School of Engineering and Applied Science – The University of Virginia, founded by Thomas Jefferson in the year 1819, believes in “developing leaders who are prepared to help in shaping the future of the nation”. The School of Engineering and Applied Science was established in 1836 and offers combined research and educational opportunities to students at undergraduate and graduate levels. The abundant research opportunities available in the school complement the curriculum, and provide a platform for young men and women to become thoughtful leaders in technology and society. The courses offered include cutting-edge research programs in computer and information science and engineering, bioengineering and nanotechnology, apart from an array of engineering disciplines.
About Interactive Supercomputing Corporation – Interactive Supercomputing Corporation (ISC) is focused on the development of Star-P, a software platform that delivers interactive parallel computing power to desktops. This platform extends the desktop simulation tools for simple user-friendly parallel computing to various computer items like SMP servers, multi-core servers, and clusters. ISC was launched in 2004 to commercialize Star-P, but the company is now working closely with Silicon Graphics Inc, a leader in high-performance computing, visualization, and storage. Customers of the company include people who solve large and complex problems that cannot be done using normal desktops.
1. U.Va. Engineering School Researchers Develop Tool for Clearer Ultrasound Images. UVa Today. University of Virginia. Last accessed July 27, 2007.
2. Facts at a Glance. University of Virginia. Last accessed July 27, 2007.
3. Company Background. Interactive Supercomputing. Last accessed July 27, 2007.
4. Sapia MA, Fox MD, Loew LM, and Schaff JC. Ultrasound Image Deconvolution Using Adaptive Inverse Filtering. Computer-Based Medical Systems, 1999. 12th IEEE Symposium. 1999:248-53. Last accessed July 27, 2007.
5. Kim YS and Ra JB. Improvement of Ultrasound Image Based on Wavelet Transform: Speckle Reduction and Edge Enhancement. Medical Imaging 2005: Image Processing. Proc of SPIE. 2005;5747:1085-92. Last accessed July 27, 2007.