Search

Ultrasound computed tomography (USCT) holds great promise for breast cancer screening. Waveform inversion-based image reconstruction methods account for higher order diffraction effects and can produce high-resolution USCT images, but are computationally demanding. Recently, a source encoding technique was combined with stochastic gradient descent to greatly reduce image reconstruction times. However, this method bundles the stochastic data fidelity term with the deterministic regularization term. This limitation can be overcome by replacing stochastic gradient descent (SGD) with a structured optimization method, such as the regularized dual averaging (RDA) method, that exploits knowledge of the composition of the cost function. In this work, the dual averaging method is combined with source encoding techniques to improve the effectiveness of regularization while maintaining the reduced reconstruction times afforded by source encoding. It is demonstrated that each iteration can be decomposed into a gradient descent step based on the data fidelity term and a proximal update step corresponding to the regularization term. Furthermore, the regularization term is never explicitly differentiated, allowing non-smooth regularization penalties to be naturally incorporated. The wave equation is solved by use of a time-domain method. The effectiveness of this approach is demonstrated through computer-simulation and experimental studies. The results suggest that the dual averaging method can produce images with less noise and comparable resolution to those obtained by use of stochastic gradient descent.

Ultrasound computed tomography (USCT) holds great promise for improving the detection and management of breast cancer. Because they are based on the acoustic wave equation, waveform inversion-based reconstruction methods can produce images that possess improved spatial resolution properties over those produced by ray-based methods. However, waveform inversion methods are computationally demanding and have not been applied widely in USCT breast imaging. In this work, source encoding concepts are employed to develop an accelerated USCT reconstruction method that circumvents the large computational burden of conventional waveform inversion methods. This method, referred to as the waveform inversion with source encoding (WISE) method, encodes the measurement data using a random encoding vector and determines an estimate of the sound speed distribution by solving a stochastic optimization problem by use of a stochastic gradient descent algorithm. Both computer-simulation and experimental phantom studies are conducted to demonstrate the use of the WISE method. The results suggest that the WISE method maintains the high spatial resolution of waveform inversion methods while significantly reducing the computational burden.

We present a multi--wavelength study of a supergiant shell within the violent interstellar medium of the nearby dwarf galaxy IC 2574, a member of the M81 group of galaxies. Neutral hydrogen (HI) observations obtained with the Very Large Array (VLA) reveal a prominent expanding supergiant HI shell which is thought to be produced by the combined effects of stellar winds and supernova explosions. It measures roughly 1000 x 500 pc in size and is expanding at about 25 km/s. The HI data suggest an age of about 1.4 x 10^6 yrs; the energy input must have been of order (2.6\pm 1) x 10^53 ergs. Massive star forming regions, as traced by H$\alpha$ emission, are situated predominantly on the rim of this HI shell. VLA radio continuum observations at 6 cm show that these star-forming regions are the main sources of radio continuum emission in this galaxy. Soft X-ray emission from within the HI hole is detected by a pointed ROSAT PSPC observation. The emission is resolved, coinciding in size and orientation with the HI shell. These spatial properties suggest that the emission is generated by an X-ray emitting plasma located within the HI shell although a contribution from X-ray binaries cannot be completely ruled out. The X-ray data are compatible with emission coming from a Raymond & Smith plasma at a temperature of about log(T[K]) = 6.8 and a density of 0.03 cm^-3., Comment: 10 pages LaTeX, 1 PS-figure, accepted by ApJ Letters

A complete H-alpha image of the Andromeda Galaxy (M31) is presented allowing the first direct measurement of the total H-alpha luminosity which is (7.3 +/- 2.4) x 10(exp 6) solar luminosity. The H-alpha emission is associated with three morphologically distinct components; a large scale star-forming ring, approximately 1.65 deg in diameter, contributing 66% of the total H-alpha emission, a bright nucleus contributing 6% of the total H-alpha emission with the remaining 28% contributed by a previously unidentified component of extended and filamentary H-alpha emission interior to the star forming ring. The correspondence between the H-alpha image and the Infrared Astronomy Satellite (IRAS) far-infrared high resolution image is striking when both are convolved to a common resolution of 105 arcsec. The close correspondence between the far-infrared and H-alpha images suggests a common origin for the two emissions. The star-forming ring contributes 70% of the far-infrared luminosity of M31. Evidence that the ring emission is energized by high mass stars includes the fact that peaks in the far-infrared emission coincide identically with H II regions in the H-alpha image. In addition, the far-infrared to H-alpha luminosity ratio within the star-forming ring is similar to what one would expect for H II regions powered by stars of spectral types ranging between O9 and B0. The origin of the filamentary H-alpha and far-infrared luminosity interior to the star-forming ring is less clear, but it is almost certainly not produced by high mass stars.