A methodology for modeling the distributions of medical images and their stochastic properties

The probabilistic distribution properties of a set of medical images are studied. It is shown that the generalized Gaussian function provides a good approximation to the distribution of AP chest radiographs. Based on this result and a goodness-of-fit test, a generalized Gaussian autoregressive model (GGAR) is proposed. Its properties and limitations are also discussed. It is expected that the GGAR model will be useful in describing the stochastic characteristics of some classes of medical images and in image data compression and other applications.     

Incorporating Actor-Critic in Monte Carlo tree search for symbolic regression

Neural Computing and Applications - Most traditional genetic programming methods that handle symbolic regression are random algorithms without memory and direction. They repeatedly search for the...     

Example-based automatic music-driven conventional dance motion synthesis

We introduce a novel method for synthesizing dance motions that follow the emotions and contents of a piece of music. Our method employs a learning-based approach to model the music to motion mapping relationship embodied in example dance motions along with those motions' accompanying background music. A key step in our method is to train a music to motion matching quality rating function through learning the music to motion mapping relationship exhibited in synchronized music and dance motion data, which were captured from professional human dance performance. To generate an optimal sequence of dance motion segments to match with a piece of music, we introduce a constraint-based dynamic programming procedure. This procedure considers both music to motion matching quality and visual smoothness of a resultant dance motion sequence. We also introduce a two-way evaluation strategy, coupled with a GPU-based implementation, through which we can execute the dynamic programming process in parallel, resulting in significant speedup. To evaluate the effectiveness of our method, we quantitatively compare the dance motions synthesized by our method with motion synthesis results by several peer methods using the motions captured from professional human dancers' performance as the gold standard. We also conducted several medium-scale user studies to explore how perceptually our dance motion synthesis method can outperform existing methods in synthesizing dance motions to match with a piece of music. These user studies produced very positive results on our music-driven dance motion synthesis experiments for several Asian dance genres, confirming the advantages of our method.     

Scaling effects on flow hydrodynamics of confined microdroplets induced by Rayleigh surface acoustic wave

This paper reports an experimental and numerical investigation on the scaling effects in the flow hydrodynamics for confined microdroplets induced by a surface acoustic wave (SAW). The characteristic parameters of the flow hydrodynamics were studied as a function of the separation height, H, between the LiNbO₃ substrate and a top glass plate, for various droplets volumes and radio-frequency powers. The ratio of the gap height to attenuation length of the SAW, H/l _SAW, is shown to be an important parameter affecting the streaming flow induced in this confined regime. The reported numerical and experimental results are in good agreement over the range examined in this study and demonstrate that, at a lower gap heights of H?≤?100?μm, a significant decrease in streaming velocity or Reynolds number is induced, with the velocity approaching zero when the gap height is decreased to ~50?μm. An increase in the gap height results in an increased streaming velocity; however, if the gap height exceeds 70?% of the SAW attenuation length, any further increase in the gap height induces a drop in the streaming velocity.     

Translational motion compensation for coronary angiogram sequences

A method of compensating for the lag of the video cameras typically used in angiographic systems is presented for use in sequences of digitized X-ray images. The lag effect is reduced by a straightforward weighted subtraction, which has the undesirable side effect of increasing noise. By superimposing several lag-corrected and appropriately shifted images, however, the signal-to-noise ratio can be restored. The algorithm uses the phase-correlation method to measure the two-dimensional shift of a mobile coronary arterial structure. Processing is confined to a rectangular area of interest (AOI), which encloses a feature of clinical significance. The differences of the phases of the Fourier transforms of two frames is computed, combined with an appropriate filter, and inverse Fourier-transformed to produce a phase-correlation image. The vector separation from the origin of image space of the peak of the phase-correlation image is the estimate of the shift of the artery's position in the second frame as compared to the first. The isolation of the AOI from the surrounding image is achieved by the application of a window and correction for any linear trend in the background intensity.     

Least squares approach in measurement-dependent filtering for selective medical images

An image-processing method called measurement-dependent filtering has been introduced to improve the SNR (signal-to-noise ratio) of selective images produced by various medical imaging systems. The basic algorithm involves the combination of the low-frequency information of the selective image with the high-frequency information of a nonselective image. A spatially variant control function modulates the amount of high frequency to be added at each point. A least-mean-square (LMS) control function formed from two basis images, namely the high-passed versions of the nonselective image (M(b)) and the selective image (S(b)), is introduced. The original algorithm is now viewed as a two-stage filtering method, including the low-pass filtering noise reduction and least squares filtering for the edge restoration. An appropriate linear transformation is used to convert the original basis images M(b) and S(b) into a new pair with orthogonal noise. This allows the implementation of the LMS and control function with practically obtainable a priori knowledge.     

Observations of slip patterns on the surface of fatigued gold using the scanning tunnelling microscope

Observations of fatigue-induced slip patterns on the surface of polycrystalline gold (+99.9%) using the scanning tunnelling microscope (STM) are presented. The samples were cycled in four-point bending between zero and 0.0025 strain for 250 and 2500 cycles. STM observations of the sample fatigued for 250 cycles revealed broad slip bands 0.38–0.69 m wide containing narrow slip bands 0.015–0.123 m wide and fine slip lines 0.006 m wide. The depth of these features are 20–57 nm for broad slip bands, 3–7 nm for narrow slip bands, and approximately 2–4 nm for fine slip lines. The sample fatigued for 2500 cycles showed similar values for the width of the slip bands and slip lines except that their depth was increased by a factor of between 2 and 4 times for the narrow slip bands and the fine slip lines. Apparent persistent slip bands (PSBs) spaced 2.4 m, 300–500 nm deep were also observed. These results demonstrate that STM is a significant new tool for observing and distinguishing various types of fatigue-induced surface slip patterns on suitably prepared samples. The vertical resolution obtained with STM is vastly superior to current SEM and TEM methods.     

Effect of nitrogen and nickel on the stability of an Fe-12Cr-23Mn austenitic steel

The effect of nitrogen and nickel on the austenite stability and the structural changes as a result of cold work of an Fe-12Cr-23Mn alloy steel has been examined by transmission electron microscopy and X-ray analysis. Prestrain was performed either at room temperature or at 200° C. Final strain was carried out at liquid nitrogen temperature. The volume fraction of strain-induced martensite decreased by 40% and 12% due to the addition of 2.43% Ni and 0.11% N, respectively, despite whether prestrain was performed or not. The mechanisms are discussed in terms of the phase, stacking faults, dislocation cells and dislocation tangles induced by prestrain and final strain.