Resolution Power Spectrum Research Articles

Usefulness of an Image-Based Noise-Reduction Technique in Cone Beam Computed Tomography for Adaptive Radiotherapy

Cone-beam computed tomography (CBCT) plays an important role in adaptive radiation therapy and can be used to observe changes in body shape, tumor size, and organ position during the treatment period. This technique can also be used to generate an image for replanning. However, performing CBCT during every image guided radiotherapy (IGRT) process increases the radiation dose. Therefore, in this study, we investigated the reduction of the CBCT exposure dose using an image-based edge-preserving noise-reduction technique. A three-dimensional cross-directional bilateral filter (3D-CDBF) which was recently developed by one of the coauthors and reported to provide a noise reduction performance better than a mode-based iterative reconstruction.Catphan was used to evaluate the performance of 3D-CDBF in CBCT. The baseline scan parameters were 120 kV, 40 mA, 40 ms, and 660 frames/360° rotation. The tube current was changed to 32, 20, and 10 mA to reduce the exposure dose. The CBCT images were reconstructed with the filtered back projection (FBP). The 3D-CDBF images were obtained by processing the FBP images. Task-based spatial resolution (task-based transfer function: TTF) and noise power spectrum (NPS) were used to evaluate image quality. Visual evaluation of the image quality was also performed by two experienced radiation oncologists and five radiologists using clinical pelvic images of two patients undergoing radiotherapy for prostate cancer. Additionally, the accuracy of the dose calculation when replanning with 3D-CDBF images was evaluated.The 50%/10% TTFs were 0.364/0.646 mm-1 using FBP and 0.362/0.645 mm-1 using 3D-CDBF, suggesting the excellent edge preservation ability of 3D-CDBF. The noise power of 3D-CDBF compared to FBP were 25.4% at 0.2 mm-1 and 17.0% at 0.5 mm-1. 3D-CDBF notably reduced noise power over the entire frequency range. Therefore, the NPS level of 10 mA became similar to that of 40 mA. The visual evaluation allowed by the 3D-CDBF images was considerably higher than that realized with other images. Additionally, dose evaluation by gamma analysis showed that 3D-CDBF had no effect on the accuracy of the dose calculation.Compared with conventional FBP, the 3D-CDBF yielded significant improvements in quantitative and qualitative image qualities. The results indicated a possible 75% dose reduction of this method, while preserving the accuracy of the dose calculation.

Effects of Image-based Noise Reduction Software on Magnetic Resonance Imaging

In magnetic resonance imaging (MRI) examinations, a trade-off is noted among acquisition time, resolution, and signal-to-noise ratio (SNR). High-resolution images are expected to improve the detection of small lesions. However, to ensure a high SNR, the imaging time must be extended. If the number of additions is reduced to shorten the imaging time, a reduction in slice thickness and in-plane resolution is necessary to ensure an adequate SNR. A combination of acceleration and denoising using deep learning has been previously reported. However, although it may be useful as a noise reduction technique onboard device, it cannot be used for general purposes. We studied the effects of a recently developed general-purpose image-based noise reduction software on MRI by measuring SNR and other parameters such as contrast, resolution, and noise power spectrum (NPS). NPS was influenced by the difference in processing mode, whereas contrast remained uninfluenced. Regarding resolution, the edge information was retained and was found to be better in iNoir 3D than in iNoir 2D. However, owing to the increased intensity of noise-reduction processing, the slope of the edge in the low-contrast area was smoothed, presenting a visually blurred impression.

Radiation Dose Reduction in CT Myocardial Perfusion Imaging Using SMART-RECON.

In this paper, a newly developed statistical model-based image reconstruction [referred to as Simultaneous Multiple Artifacts Reduction in Tomographic RECONstruction (SMART-RECON)] is applied to low dose computer tomography (CT) myocardial perfusion imaging (CT-MPI). This method uses the nuclear norm of the spatial-temporal image matrix of the CT-MPI images as a regularizer, rather than a conventional spatial regularizer that incorporates image smoothness, edge preservation, or spatial sparsity into the reconstruction. In addition to providing the needed noise reduction for low-dose CT-MPI, SMART-RECON provides images with spatial resolution and noise power spectrum (NPS) properties, which are independent of contrast and dose levels. Both numerical simulations and in vivo animal studies were performed to validate the proposed method. In these studies, it was found that: 1) quantitative accuracy of perfusion maps in CT-MPI was well maintained for radiation dose level as low as 10 mAs per image frame, compared with the reference standard of 200 mAs for conventional filtered backprojection; 2) flow-occluded myocardium in the porcine heart was well delineated by SMART-RECON at 10 mAs per frame when compared with model-based image reconstruction using spatial total variation (TV) as the regularizer (referred to as TV-SIR) or spatial-temporal TV (ST-TV-SIR); the CT-MPI results were confirmed with positron-emission tomography imaging; 3) image sharpness in SMART-RECON images was nearly independent of image contrast level and radiation dose level, in stark contrast to TV-SIR and ST-TV-SIR, which displayed a strong dependence on both image contrast and radiation dose levels; and 4) the structure of the dose-normalized NPS for the SMART-RECON method did not depend on dose, while the TV-SIR and ST-TV-SIR NPS structure was dose-dependent.

Effects of Image Quality on the Fundamental Limits of Image Registration Accuracy.

For image-guided procedures, the imaging task is often tied to the registration of intraoperative and preoperative images to a common coordinate system. While the accuracy of this registration is a vital factor in system performance, there is a relatively little work that relates registration accuracy to image quality factors, such as dose, noise, and spatial resolution. To create a theoretical model for such a relationship, we present a Fisher information approach to analyze registration performance in explicit dependence on the underlying image quality factors of image noise, spatial resolution, and signal power spectrum. The model yields analysis of the Cramer-Rao lower bound (CRLB), in registration accuracy as a function of factors governing image quality. Experiments were performed in simulation of computed tomography low-contrast soft tissue images and high-contrast bone (head and neck) images to compare the measured accuracy [root mean squared error (RMSE) of the estimated transformations] with the theoretical lower bound. Analysis of the CRLB reveals that registration performance is closely related to the signal-to-noise ratio of the cross-correlation space. While the lower bound is optimistic, it exhibits consistent trends with experimental findings and yields a method for comparing the performance of various registration methods and similarity metrics. Further analysis validated a method for determining optimal post-processing (image filtering) for registration. Two figures of merit (CRLB and RMSE) are presented that unify models of image quality with registration performance, providing an important guide to optimizing intraoperative imaging with respect to the task of registration.

Noise Suppression for Dual-Energy CT Through Entropy Minimization.

In dual energy CT (DECT), noise amplification during signal decomposition significantly limits the utility of basis material images. Since clinically relevant objects typically contain a limited number of different materials, we propose an Image-domain Decomposition method through Entropy Minimization (IDEM) for noise suppression in DECT. Pixels of decomposed images are first linearly transformed into 2D clusters of data points, which are highly asymmetric due to strong signal correlation. An optimal axis is identified in the 2D space via numerical search such that the projection of data clusters onto the axis has minimum entropy. Noise suppression is performed on each image pixel by estimating the center-of-mass value of each data cluster along the direction perpendicular to the projection axis. The IDEM method is distinct from other noise suppression techniques in that it does not suppress pixel noise by reducing spatial variation between neighboring pixels. As supported by studies on Catphan©600 and anthropomorphic head phantoms, this feature endows our algorithm with a unique capability of reducing noise standard deviation on DECT decomposed images by approximately one order of magnitude while preserving spatial resolution and image noise power spectra (NPS). Compared with a filtering method and recently developed iterative method at the same level of noise suppression, the IDEM algorithm obtains high-resolution images with less artifacts. It also maintains accuracy of electron density measurements with less than 2% bias error. The IDEM method effectively suppresses noise of DECT for quantitative use, with appealing features on preservation of image spatial resolution and NPS.

Radio access technology recognition by classification of low temporal resolution power spectrum measurements

AbstractIn this paper, a machine learning approach is proposed for automatic recognition of the radio access technology (RAT) based on supervised classification of spectrum power measurements in a cognitive radio context. Recognition of the active local RATs allows the cognitive terminal to make informed decisions on which wireless technology to camp, and to adapt its transmission to the wireless environment. To this end, supervised classification of signals obtained from wide band spectrum power measurements with low temporal resolution is carried out. As a first step, generic features are extracted from the power spectrum measurements. Then several supervised classification algorithms, namely the support vector machines (SVMs), k‐nearest neighbors (kNN) and C4.5 are tested. The effects of reducing the number of features on classification accuracy is also tested through feature selection methods. Finally, the impact of the signal duration on classification accuracy is investigated. The obtained results demonstrate that the tested supervised classification algorithms, combined with simple feature extraction and selection methods, are good candidates for RAT recognition from low temporal resolution spectrum power measurements. Copyright © 2009 John Wiley & Sons, Ltd.

TU-A-201B-02: 20-Fold Dose Reduction Using a Gradient Adaptive Bilateral Filter: Demonstration Using in Vivo Animal Perfusion CT

Purpose: To demonstrate the dose saving potential of a fast and practical algorithm based on a novel Gradient Adaptive Bilateral filtration (GABI) algorithm, applied to in vivo extremely low dose animal perfusion CT scans. Methods: the GABI algorithm is mainly applied in the time dimension to avoid losing spatial resolution and noise power spectra characteristics (texture of CT images). It expands the bilateral filter algorithm by non-locally adapting its strength based on a temporal gradient applied to the image dataset, and hence preserving contrast fidelity over time. We tested the GABI on two renal perfusion animal experiments which used a reference dose of 160 mAs (full dose FD), and 16 mAs and 8 mAs (1/10th and 1/20th of the radiation dose). Results: we demonstrated that the GABI algorithm preserves the temporal fidelity, as measured by the time attenuation curves (TACs), while dramatically reducing image noise from very low dose acquisitions, keeping comparable spatial resolution, and preserving CT noise texture. The background SNR for the FD pig A, FD pig B, 1/10th dose pig A, 1/20th dose pig B, 1/10th pig A + GABI, 1/20th pig B + GABI were 4.5, 4.4, 1.4, 0.85, 3.8 and 2.9 respectively. Higher background SNR improvements are possible, however we were constrained to keep very high fidelity to the TACs and hence perfusion parameters. Conclusions: the GABI algorithm provides a practical framework to highly reduce the dose in CT time series studies. The method is fast since it is image-space based and non-iterative. We demonstrated in real animal data, that in high SNR tasks as renal perfusion the gain in dose reduction can be as high as 20-fold with no practical detriment of perfusion information and image quality comparable to a full dose image acquisition. Further investigation for parameter optimization is guaranteed.

WE-D-303A-05: How to Measure Fluoroscopic “dose Efficiency”: The Spatio-Temporal Detective Quantum Efficiency

Purpose: To measure the “dose efficiency” of fluoroscopic systems on a quantitative and absolute scale by developing a spatio-temporal detective quantum efficiency (DQE) metric of performance. Method and Materials: We developed the first comprehensive spatio-temporal approach to fluoroscopic system performance through the development of a spatio-temporal DQE. It is defined in terms of the presampling spatio-temporal modulation transfer function (MTF), describing resolution, and spatio-temporal noise power spectrum (NPS), describing noise. We developed what we call the “semi-transparent moving slanted-edge method” to measure the temporal component of the MTF, which uses a small-signal approach to temporal resolution. This method overcomes temporal non-linearity problems associated with differences between lag build-up and decay responses of CsI-based systems. A three-dimensional (3D) approach to fluoroscopic noise, in which 3D spatio-temporal ROIs are selected in space and time, was used to calculate the spatio-temporal NPS. Spatio-temporal DQE values of a bench-top x-ray image intensifier fluoroscopic system were calculated, illustrating use of the metric and gaining insight into fluoroscopic temporal dose efficiency. Results: Non-ideal temporal performance was noted in the test system at temporal frequencies > 5 Hz, and there was a ∼50% drop in DQE values by 15 Hz, analogous to decreases in DQE values commonly seen at large spatial frequencies. This is the first time that fluoroscopic temporal performance has been quantized using a DQE metric, and the non-ideal system temporal performance may be due to temporal noise aliasing and the statistical nature of lag. Conclusion: The spatio-temporal DQE describes fluoroscopic system performance and dose efficiency in both space and time, providing temporal information that can not be obtained using lag-corrected or other spatial approaches. The ∼50% decrease in performance by 15 Hz may suggest the presence of temporal quantum sinks.

A Simulation Study on Spatial Resolution and Noise Power Spectra of a URA-based Multi-hole Collimator in a Small Gamma Camera

Presently the gamma scintillation camera is widely used in various industrial, environmental and medical diagnostic fields. Two major performance parameters, spatial resolution and noise, are primarily determined by the collimator. The imaging performance of the simple collimator, single pin-hole, and a coded-aperture collimator, uniformly redundant array (URA), are analyzed in this study. Though parallel-hole collimators are used in some medical applications, in principle its performance is equivalent to that of a single-hole collimator and moreover its use is limited to the near-field application only. A coded aperture imaging (CAI) techniques have been proposed for gamma-ray imaging especially for far-field applications such as the astrophysics study or environmental monitoring in order to overcome these limitations of a pin-hole or a parallel-hole collimator.In this study, the simulated gamma images using a monte carlo code, MCNP, were acquired with two different energy windows and two-type and four-different collimators. For the evaluation of spatial resolution and noise power spectra, modulation transfer function (MTF) and normalized noise power spectra (NNPS) were computed for all images. In the result, we found that MTF and NNPS can reflect the different properties of different gamma camera systems.

SU‐FF‐I‐97: Feasibility of Low Dose X‐Ray Contrast Enhanced Digital Mammography with Gold Nanoparticles

Purpose: To study the feasibility of low dose x‐ray contrast enhanced digital mammography (CEDM) with gold nanoparticles as a contrast agent. Method and Materials: Contrast enhanced digital mammography (CEDM) with iodine based agents is currently being explored. However, novel materials such as gold nanoparticles seem very promising for CEDM because of their high x‐ray absorption in the mammographic energy range. Computer generated volumetric breast compositions of 50% and 75% mean glandular fractions and thickness of 5 and 7.5 cm were analytically ray‐traced at various mono‐energetic conditions and spectrally combined with simulated power‐law noise to create 2D projection images with mammogram‐like texture at 100 micron pixel resolution. The projection images were modified in accordance with the system resolution and noise power spectrum of a clinical full‐field digital mammography system (Senographe 2000D, GE Healthcare, WI) and used to generate image data sets. A 6 mm diameter spherical lesion with an inherent glandular fraction of 65% and 100% was used for the 50% and 75% glandular breast conditions respectively. Lesions with contrast agent concentrations of 0.5, 1, 2 and 4 mg/cc were generated. A 26 kVp, Mo/Mo and 31 kVp, Rh/Rh x‐ray spectra were used to generate full‐dose images (∼2 mGy mean glandular dose) for the 5 and 7.5 cm‐thick breasts respectively. A Laguerre‐Gauss Channelized Hotelling observer (LG‐CHO) was implemented to compute the signal‐to‐noise ratio (SNR). Results: Good contrast visualization was observed at ∼0.13 mGy per image frame with CEDM. Contrast agent concentrations between 0.5–2 mg/cc in the lesion resulted in much higher SNRs compared to full dose mammography. Conclusion: This work suggests favorable properties for gold nanoparticles as an x‐ray contrast agent for CEDM. This could potentially benefit high‐risk patient groups such as women with dense breasts. Other potential applications include progress monitoring in neo‐adjuvant treatment regimens.

Distribution of focal lesions in the chinchilla organ of Corti following exposure to a 4-kHz or a 0.5-kHz octave band of noise

An octave band of noise (OBN) delivers fairly uniform acoustic energy over a specific range of frequencies. Above and below this range, energy is at least 30 dB SPL less than that within the OBN. When the ear is exposed to an OBN, hair-cell loss often occurs outside the octave band. The frequency location of hair-cell loss is evident when the percent distance from the apex of focal lesions is analyzed. Focal lesions involve substantial loss of outer hair cells (OHCs) only, inner hair cells (IHCs) only, or both OHCs and IHCs (i.e., combined lesions) in a specific region of the organ of Corti (OC). Data sets were assembled from our permanent collection of noise-exposed chinchillas as follows: (1) the sum of exposure duration and recovery time was less than or equal to 11 d; (2) the exposure level was less than or equal to 108 dB SPL; and (3) focal lesions were less than 1.5mm in length. The data sets included a variety of exposures ranging from high-level, short duration to moderate-level, moderate duration. The center of each focal lesion was expressed as percent distance from the OC apex. Means, standard deviations and medians were calculated for focal-lesion size resulting from exposure to a 4-kHz or a 0.5-kHz OBN. Histograms were then constructed from the percent-location data using 2.0% bins. For the 4-kHz OBN, 5% of the lesions were in the apical half of the OC and 95% were in the basal half. The mean lesion size was 1.68% of total OC length for OHC and combined focal lesions and 0.42% for IHC focal lesions. Most OHC and combined lesions occurred in the 5-7-kHz region, at and just above the upper edge of the OBN. Clusters of lesions were also found around 8 and 12 kHz. A cluster was present at and just below the lower edge of the OBN, as well as another in the 1.5-kHz region. For the 0.5-kHz OBN, 34% of the lesions were in the apical half of the OC and 66% were in the basal half. The mean lesion size was 0.93% for OHC and combined focal lesions and 0.32% for IHC focal lesions. OHC and combined focal-lesion distribution showed clusters at 0.25, 0.75 and 1.5 kHz in the apical half of the OC. In the basal half, the distribution of focal lesions was similar to that seen with the 4-kHz OBN (r=0.54). With both OBNs, most IHC focal lesions occurred in the basal half of the OC. High resolution power spectrum analysis of each OBN and non-invasive tests for harmonics and distortion products in a chinchilla were performed to look for exposure energy above and below the OBN. No energy was found that could explain the OC damage.

The Periodicities of Solar X-ray Flares and Coronal Mass ejections during Solar Cycle 23

The Events of energetic particles from solar X-ray flares and shock waves have been studied. The data were taken from the National Geophysical data center (NGDC) in Boulder, Colorado, USA, Where the data were taken during the solar cycle 23rd. The 23rd cycle is the present one, that started in April 1996, and its maximum was in May 2001, and it will be decayed during year 2007. Power spectrum methods have been applied for analysis of the data given, to find the short and intermediate periodicities. The periodicity around 14 days has appeared in this analysis. This has important implications for understanding and predicting the effects of solar activity on the Earth and on the earth’s atmosphere. If a Coronal Mass Ejections (CMEs) hit the Earth, it can excite a geomagnetic storm. Large geomagnetic storms, among other things, can cause electrical power which can damage satellite communications. In space CME typically drive shock waves that produce energetic particles that can damage both electronic equipment and astronauts that are outside the protection of the Earth’s magnetic field. So, the prediction of the high energetic particle events is of vital importance for space navigation and airline disasters. 1. Method and Analysis The data of events occurred during solar cycle 23rd, were taken from the National Geophysical data center (NGDC) in Boulder, Colorado, USA. The method of power spectrum analysis of the data has been used for all data analysis, where the squared amplitudes [aν ] are taken as the power of the frequency ν, as follows: [aν ] = ( ∑N−1 i=0 r(i) cos(2πi) )2 + ( ∑N−1 i=0 r(i) sin(2πi) )2 N2 Where N is the total number of X-ray flares events, equal to 6483 events in solar Xray calculations of power spectrum. By using hamming function the data have been smoothed, where the resolution power spectrum given as: rn = 0.23rn−1 + 0.54rn + 0.23rn+1 For more details about the method of analysis, see for example, Hady A. (2002), and Rieger et al (1984).The relative error values are calculated according to:

Computer modeling of loudspeaker arrays in rooms

Loudspeakers present a special challenge to computational modeling of rooms. When modeling a collection of noncorrelated sound sources, such as a group of musicians, coarse resolution power spectrum and directivities are sufficient. In contrast, a typical loudspeaker array consists of many speakers driven with the same signal, and are therefore almost completely correlated. This can lead to a quite complicated, but stable, pattern of spatial nulls and lobes which depends sensitively on frequency. It has been shown that, to model these interactions accurately, one must have loudspeaker data with 1 deg spatial resolution, 1/24 octave frequency resolution including phase. It will be shown that computer models at such a high resolution can in fact inform design decisions of loudspeaker arrays.

Long-Period Chromospheric Oscillations in Network Bright Points

The spatial variation of chromospheric oscillations in network bright points (NBPs) is studied using high-resolution observations in Ca II K3. Light curves and hence power spectra were created by isolating distinct regions of the NBP via a simple intensity thresholding technique. Using this technique, it was possible to identify peaks in the power spectra with particular spatial positions within the NBPs. In particular, long-period waves with periods of 4-15 minutes (1-4 mHz) were found in the central portions of each NBP, indicating that these waves are certainly not acoustic but possibly due to magnetoacoustic or magnetogravity wave modes. We also show that spatially averaged or low spatial resolution power spectra can lead to an inability to detect such long-period waves.

A method for extracting maximum resolution power spectra from microwave sky maps

A method for extracting maximal resolution power spectra from microwave sky maps is presented and applied to the 2 year COBE data, yielding a power spectrum that is consistent with a standard n=1, Q=20 micro-Kelvin model. By using weight functions that fall off smoothly near the galactic cut, it is found that the spectral resolution \Delta l can be more than doubled at l=15 and more than tripled at l=20 compared to simply using galaxy-cut spherical harmonics. For a future high-resolution experiment with reasonable sky coverage, the resolution around the CDM Doppler peaks would be enhanced by a factor of about 100, down to \Delta l\approx 1, thus allowing spectral features such as the locations of the peaks to be determined with great accuracy. The reason that the improvement is so large is basically that functions with a sharp edge at the galaxy cut exhibit considerable "ringing" in the Fourier domain, whereas smooth functions do not. The method presented here is applicable to any survey geometry, chopping strategy and exposure pattern whatsoever. The so called signal-to-noise eigenfunction technique is found to be a special case, corresponding to ignoring the width of the window functions.

A linear programming approach to the estimation of the power spectra of harmonic processes

A new approach to the estimation of high resolution power spectra of harmonic processes is presented. This approach uses linear programming (LIP) to exploit the spiky character of the L 1 norm. Results for the LIP power spectral estimate are contrasted with those using various formulations of the maximum entropy algorithm.

Solar and lunar daily geomagnetic variations at Dourbes

Abstract Spectral analysis of the Dourbes H component hourly data from the period 1960–1978 revealed the existence of a number of minor terms, in addition to the main solar and lunar peaks. The relative amplitudes of oscillations in the geomagnetic spectrum are unrelated with those predicted through lunar tide theory. The minor terms agree more closely with the 27-day amplitude modulation mechanism. A high frequency resolution power spectrum clearly shows the splitting of the solar diurnal and semi-diurnal line, and even of the lunar semi-diurnal line by the annual variation and its harmonics. The correlation between the amplitude of the M2 wave and the mean sunspot number is of no significance.

High resolution geomagnetic spectra in Indian region: Lunar tides

Abstract A search for major and minor lunar geomagnetic tides in the Indian region is made from high resolution power spectra, obtained by the Maximum Entropy Method (MEM), at two low latitude stations. The results indicate, on the basis of frequency comparison, that only near the dip equator is there clear evidence for the lunar tidal mechanism being the dominant factor over the amplitude modulation mechanism. Amplitudes of the lunar terms near 1 and 2 c day −1 show equatorial enhancement with greater enhancement for the semidiurnal terms. The lunar semidiurnal term, M 2 , shows the largest enhancement.

Variation of the solar atmospheric rotation over the 11 year cycle

The synodic rotation period and power spectra of solar microwave sources are investigated using accurate data in the interval 1956 to 1970. The variation of the approximate 27 day period is obtained over a complete solar cycle and is thought to be a result of the latitude change over the solar cycle of the origins of the radio emissions. High resolution power spectra have also been obtained and revealed the existence of a double peaked line near 160 day period. This line is attributed to changes in either the Eartn's heliographic latitudes or the Earth's inclination to the Earth-Sun line.

A 1024−channel digital correlator

A 1024−channel digital correlator designed for spectral analysis of radio astronomy signals is described. The correlator has a maximum sampling rate of 20 MHz, allowing it to analyze noise signals with bandwidths up to 10 MHz. Other bandwidths of 5, 2, 1, 0.5, 0.2, 0.1, and 0.05 MHz are provided. Quantization mode can be either one−bit or two−bit. Either autocorrelation or cross−correlation functions can be generated. These are Fourier transformed on−line by a PDP−9 computer to generate high resolution power spectra. Subdivision of the correlator into four independent 256−channel units is also possible. Software used with the correlator is briefly described, and some observational results presented.