Spectral-based Techniques Research Articles

Spectral-based estimation of chlorophyll content and determination of background interference mechanisms in low-coverage rice

The chlorophyll content is a vital indicator of rice growth and nutritional status. However, estimating the rice chlorophyll content using spectral-based techniques at the early tillering stage is challenging because of background interference. Using the energy conservation principle, this study explained the spectral variation and background interference mechanisms of clear, muddy, and green algae-covered backgrounds. We developed mathematical interference models for the three types of backgrounds and determined their interference degree and influence mode. We developed rice chlorophyll content estimation models for unclassified and classified (clear, muddy, and green algae-covered) backgrounds using 12 preprocessing, four wavelength selection, and three modeling methods, and we explored the importance of background classification. Moreover, we found that the optimal chlorophyll content estimation model for the clear background was SS+UVE+CNN, with R2 and RMSE values of 0.786 and 13.191 in the training set and 0.741 and 15.327 in the test set, respectively; that for the muddy background was MSC+GA+CNN, with R2 and RMSE values of 0.914 and 10.425 in the training set and 0.660 and 16.844 in the test set, respectively; and that for the green algae-covered background was DC+GA+CNN, with R2 and RMSE values of 0.904 and 9.111 in the training set and 0.688 and 17.694 in the test set, respectively. Our study could provide valuable insights into reducing and correcting background interference during proximal remote sensing data collection.

Cancer characterization using light backscattering spectroscopy and quantitative ultrasound: an ex vivo study on sarcoma subtypes

Histological analysis is the gold standard method for cancer diagnosis. However, it is prone to subjectivity and sampling bias. In response to these limitations, we introduce a quantitative bimodal approach that aims to provide non-invasive guidance towards suspicious regions. Light backscattering spectroscopy and quantitative ultrasound techniques were combined to characterize two different bone tumor types from animal models: chondrosarcomas and osteosarcomas. Two different cell lines were used to induce osteosarcoma growth. Histological analyses were conducted to serve as references. Three ultrasound parameters and intensities of the light reflectance profiles showed significant differences between chondrosarcomas and osteosarcomas at the 5% level. Likewise, variations in the same biomarkers were reported for the two types of osteosarcoma, despite their similar morphology observed in the histological examinations. These observations show the sensitivity of our techniques in probing fine tissue properties. Secondly, the ultrasound spectral-based technique identified the mean size of chondrosarcoma cells and nuclei with relative errors of about 22% and 9% respectively. The optical equivalent technique correctly extracted scatterer size distributions that encompass nuclei and cells for chondrosarcomas and osteosarcomas (R^2 = 0.80 and R^2 = 0.73 respectively). The optical scattering contributions of nuclei were estimated at 52% for the chondrosarcomas and 69% for the osteosarcomas, probably indicating the abundant and the absent extracellular matrix respectively. Thus, the ultrasound and the optical methods brought complementary parameters. They successfully estimated morphological parameters at the cellular and the nuclear scales, making our bimodal technique promising for tumor characterization.

Macroalgae monitoring from satellite optical images using Context-sensitive level set (CSLS) model

Automatically detecting macroalgae from optical remote sensing images has been a long-standing problem in ocean monitoring. Existing operational spectral-based techniques tend to assign a threshold to an indicator such as Normalized Difference Vegetation Index (NDVI) to determine whether or not macroalgae exists in a pixel. This thresholding scheme relies on individual pixel features and cannot address the environment variability among different images or even across one single image. To address these limitations, we develop the unsupervised classification method taking the ecological indicator of NDVI as input by exploiting a region-based level set driven by context-sensitive energy function for macroalgae detection from optical remote sensing images. Quantitative and qualitative evaluations based on satellite optical images with various imaging time and spatial resolutions (16–250 m) from GF-1, GF-4, Landsat-8, and Aqua demonstrate the effectiveness and robustness for Ulva prolifera detection. The model is proved to be superior to the traditional NDVI thresholding method and state-of-the-art level set methods in terms of higher accuracy (Kappa coefficients 0.87–0.92) and automation (without artificial threshold/parameter adjustment). Moreover, the model has the advantage of accommodating complex environmental conditions (such as different water turbidity, thin clouds and sun glint), and can be utilized for detection of Sargassum bloom. The model could facilitate the accurate extraction of macroalgae information and thus serve disaster prevention and control. It may also be used to assist high-accuracy training data preparation for supervised classification and detect the objects with obvious texture and spectral characteristics (such as oil spills) based on customized bands combination.

A Review of Key Techniques for in Ovo Sexing of Chicken Eggs

The identification of chicken sex before hatching is an important problem in large-scale breeding applications in the poultry industry. This paper systematically reviews the key techniques for in ovo sexing of chicken eggs before hatching and presents recent research on molecular-based, spectral-based, acoustic-based, morphology-based, and volatile organic compound (VOC)-based technologies. Molecular-based methods are standard techniques for accurate sexing but require perforations by skilled technicians in certified laboratories to extract egg contents. Spectral-based techniques show great potential as noninvasive methods but require complex data processing and modeling. Acoustic-based techniques are sensitive to environmental noise. Morphology-based studies on the outer shape of the eggshell and distribution of blood vessels provide novel methods for in ovo sexing of chicken eggs. However, they face challenges such as the color, thickness, and smoothness of the eggshell. VOC profiling of chicken eggs allows sexing in the early stages of incubation; however, the VOC composition may be influenced by species or feed, and more research is required to explore potential applications. In addition, recent breakthroughs on in ovo chicken egg sexing are discussed. Physiological changes in chicken eggs during the whole incubation period have been well studied using metabolism and phenotype tools to enhance mechanism recognition. Furthermore, various sensing techniques, from the X-ray to terahertz range, and deep learning algorithms have been employed for data acquisition, processing, mining, and modeling to capture and analyze key features. Finally, commercialization and practical applications are discussed. This study provides a reference for in ovo sexing of chicken eggs before hatching in the poultry industry.

Comparison of two spectral-based techniques for estimating the attenuation coefficient from human cervix

This study uses in vivo radiofrequency ultrasound data acquired from human cervices to compare two commonly used spectral-based techniques for estimating the ultrasonic attenuation coefficient (AC): the spectral difference (SD) and the spectral log difference (SLD) techniques. The AC is a fundamental quantitative ultrasound (QUS) parameter useful for tissue characterization to improve diagnostics, e.g., cervix characterization to predict preterm birth. The selection of appropriate AC techniques is a critical step for QUS tissue characterization. The advantages and disadvantages of various AC estimation techniques have been studied using physical phantoms and computational simulations. However, the heterogeneous nature of real tissue cannot be fully simulated with phantoms and computations. In this study, the SD and SLD techniques were evaluated using human cervices from 214 pregnant women (enrollment still ongoing). Each participant underwent 1-2 cervical QUS exams during each visit. In each exam, 10 acquisitions of radiofrequency ultrasound and one acquisition of a physical phantom were made using a Siemens MC9-4 transvaginal ultrasound transducer (center frequency: 5.25 MHz) with a Siemens S2000 ultrasound system. Preliminary analysis yielded correlated AC estimates (Pearson’s r = 0.66; <0.001) between the two AC techniques and better precision (e.g., inter-sonographer reproducibility) with the SLD technique. [No. R01HD089935.]

A spatialized classification approach for land cover mapping using hyperspatial imagery

Maps of classified surface features are a key output from remote sensing. Conventional methods of pixel-based classification label each pixel independently by considering only a pixel's spectral properties. While these purely spectral-based techniques may be applicable to many medium and coarse-scale remote sensing analyses, they may become less appropriate when applied to high spatial resolution imagery in which the pixels are smaller than the objects to be classified. At this scale, there is often higher intra-class spectral heterogeneity than inter-class spectral heterogeneity, leading to difficulties in using purely spectral-based classifications. A solution to these issues is to use not only a pixel's spectral characteristics but also its spatial characteristics. In this study, we develop a generalizable "spatialized" classification approach for high spatial resolution image classification. We apply the proposed approach to map vegetation growth forms such as trees, shrubs, and herbs in a forested ecosystem in the Sierra Nevada Mountains. Our results found that the spatialized classification approach outperformed spectral-only approaches for all cover classes examined, with the largest improvements being in discriminating vegetation classes.

Geometric classification of brain network dynamics via conic derivative discriminants

BackgroundOver the past decade, pattern decoding techniques have granted neuroscientists improved anatomical specificity in mapping neural representations associated with function and cognition. Dynamical patterns are of particular interest, as evidenced by the proliferation and success of frequency domain methods that reveal structured spatiotemporal rhythmic brain activity. One drawback of such approaches, however, is the need to estimate spectral power, which limits the temporal resolution of classification. New methodWe propose an alternative method that enables classification of dynamical patterns with high temporal fidelity. The key feature of the method is a conversion of time-series data into temporal derivatives. By doing so, dynamically-coded information may be revealed in terms of geometric patterns in the phase space of the derivative signal. ResultsWe derive a geometric classifier for this problem which simplifies into a straightforward calculation in terms of covariances. We demonstrate the relative advantages and disadvantages of the technique with simulated data and benchmark its performance with an EEG dataset of covert spatial attention. We reveal the timecourse of covert spatial attention and, by mapping the classifier weights anatomically, its retinotopic organization. Comparison with existing methodWe especially highlight the ability of the method to provide strong group-level classification performance compared to existing benchmarks, while providing information that is complementary with classical spectral-based techniques. The robustness and sensitivity of the method to noise is also examined relative to spectral-based techniques. ConclusionThe proposed classification technique enables decoding of dynamic patterns with high temporal resolution, performs favorably to benchmark methods, and facilitates anatomical inference.

Regularized Spectral Log Difference Technique for Ultrasonic Attenuation Imaging.

The attenuation coefficient slope (ACS) has the potential to be used for tissue characterization and as a diagnostic ultrasound tool, hence complementing B-mode images. The ACS can be valuable for the estimation of other ultrasound parameters such as the backscatter coefficient. There is a well-known tradeoff between the precision of the estimated ACS values and the data block size used in the spectral-based techniques such as the spectral-log difference (SLD). This tradeoff limits the practical usefulness of the spectral-based attenuation imaging techniques. In this paper, the regularized SLD (RSLD) technique is presented in detail, and evaluated with simulations and experiments with physical phantoms, ex vivo and in vivo. The RSLD technique allowed decreasing estimation variance when using small data block sizes, i.e., fivefold reduction in the standard deviation of percentage error when using data block sizes larger than and more than a tenfold reduction when using data blocks. The precision improvement was obtained without sacrificing estimation accuracy (i.e., estimation bias improved in 70% of the cases by 10% of the ground truth-value on average while degraded in 30% of the cases by 3% of the ground truth-value on average). The improvements in precision allowed for better differentiation of inclusions especially when using small data blocks (i.e., smaller than ) where the contrast-to-noise ratio improved by an order of magnitude on average. The results suggest that the RSLD allows for the reconstruction of attenuation coefficient images with an improved tradeoff between spatial resolution and estimation precision.

Emerging quantitative ultrasound applications: From diagnosing disease to monitoring of therapy

Ultrasound has traditionally been considered an imaging modality characterized by high spatial resolution but low contrast. Conventional ultrasonic imaging may be sensitive to the detection of anomalous tissue features, but the ability to classify these tissue features often lacks specificity. However, recent advances in quantitative ultrasound (QUS) have emerged which can improve diagnostics by providing new sources of image contrast and specific numbers tied to tissue state. QUS imaging can encompass a wide variety of techniques including spectral-based parameterization, elastography, flow estimation, and envelope statistics. Spectral-based techniques include the estimation of the backscatter coefficient, estimation of attenuation, and estimation of scatterer properties such as the correlation length associated with an effective scatterer size and the acoustic concentration of scatterers. Envelope statistics include the estimation of the number density of scatterers and quantification of coherent to incoherent signals produced from the tissue. In this talk, successful applications of QUS will be discussed demonstrating the ability of spectral-based QUS and envelope statistics approaches to improve medical diagnostics including cancer detection and classification of solid tumors in the breast or thyroids, grading of fatty liver disease and monitoring and assessment of therapy in solid tumors.

Spectral algorithm for non-destructive damage localisation: Application to an ancient masonry arch model

Structural monitoring and vibration-based damage identification methods are fundamental tools for condition assessment and early-stage damage identification, especially when dealing with the conservation of historical constructions and the maintenance of strategic civil structures. However, although the substantial advances in the field, several issues must still be addressed to broaden the application range of such tools and to assert their reliability. This study deals with the experimental validation of a novel method for non-destructive damage identification purposes. This method is based on the use of spectral output signals and has been recently validated by the authors through a numerical simulation. After a brief insight into the basic principles of the proposed approach, the spectral-based technique is applied to identify the experimental damage induced on a masonry arch through statically increasing loading. Once the direct and cross spectral density functions of the nodal response processes are estimated, the system's output power spectrum matrix is built and decomposed in eigenvalues and eigenvectors. The present study points out how the extracted spectral eigenparameters contribute to the damage analysis allowing to detect the occurrence of damage and to locate the target points where the cracks appear during the experimental tests. The sensitivity of the spectral formulation to the level of noise in the modal data is investigated and discussed. As a final evaluation criterion, the results from the spectrum-driven method are compared with the ones obtained from existing non-model based damage identification methods.

Review of Quantitative Ultrasound: Envelope Statistics and Backscatter Coefficient Imaging and Contributions to Diagnostic Ultrasound.

Conventional medical imaging technologies, including ultrasound, have continued to improve over the years. For example, in oncology, medical imaging is characterized by high sensitivity, i.e., the ability to detect anomalous tissue features, but the ability to classify these tissue features from images often lacks specificity. As a result, a large number of biopsies of tissues with suspicious image findings are performed each year with a vast majority of these biopsies resulting in a negative finding. To improve specificity of cancer imaging, quantitative imaging techniques can play an important role. Conventional ultrasound B-mode imaging is mainly qualitative in nature. However, quantitative ultrasound (QUS) imaging can provide specific numbers related to tissue features that can increase the specificity of image findings leading to improvements in diagnostic ultrasound. QUS imaging can encompass a wide variety of techniques including spectral-based parameterization, elastography, shear wave imaging, flow estimation, and envelope statistics. Currently, spectral-based parameterization and envelope statistics are not available on most conventional clinical ultrasound machines. However, in recent years, QUS techniques involving spectral-based parameterization and envelope statistics have demonstrated success in many applications, providing additional diagnostic capabilities. Spectral-based techniques include the estimation of the backscatter coefficient (BSC), estimation of attenuation, and estimation of scatterer properties such as the correlation length associated with an effective scatterer diameter (ESD) and the effective acoustic concentration (EAC) of scatterers. Envelope statistics include the estimation of the number density of scatterers and quantification of coherent to incoherent signals produced from the tissue. Challenges for clinical application include correctly accounting for attenuation effects and transmission losses and implementation of QUS on clinical devices. Successful clinical and preclinical applications demonstrating the ability of QUS to improve medical diagnostics include characterization of the myocardium during the cardiac cycle, cancer detection, classification of solid tumors and lymph nodes, detection and quantification of fatty liver disease, and monitoring and assessment of therapy.

Experimental Application of Ultrafast Imaging to Spectral Tissue Characterization

Ultrasound ultrafast imaging (UI) allows acquisition of thousands of frames per second with a sustained image quality at any depth in the field of view. Therefore, it would be ideally suited to obtain good statistical sampling of fast-moving tissues using spectral-based techniques to derive the backscatter coefficient (BSC) and associated quantitative parameters. In UI, an image is formed by insonifying the medium with plane waves steered at different angles, beamforming them and compounding the resulting radiofrequency images. We aimed at validating, experimentally, the effect of these beamforming protocols on the BSC, under both isotropic and anisotropic conditions. Using UI techniques with a linear array transducer (5–14 MHz), we estimated the BSCs of tissue-mimicking phantoms and flowing porcine blood at depths up to 35 mm with a frame rate reaching 514 Hz. UI-based data were compared with those obtained using single-element transducers and conventional focusing imaging. Results revealed that UI compounded images can produce valid estimates of BSCs and effective scatterer size (errors less than 2.2 ± 0.8 and 0.26 ± 0.2 dB for blood and phantom experiments, respectively). This work also describes the use of pre-compounded UI images (i.e., steered images) to assess the angular dependency of circulating red blood cells. We have concluded that UI data sets can be used for BSC spectral tissue analysis and anisotropy characterization.

Nonlinear stability, bifurcation and vortical patterns in three-dimensional granular plane Couette flow

AbstractThe effects of three-dimensional (3D) perturbations, having wave-like modulations along both the streamwise and spanwise/vorticity directions, on the nonlinear states of five types of linear instability modes, the nature of their bifurcations and the resulting nonlinear patterns are analysed for granular plane Couette flow using an order-parameter theory which is an extension of our previous work on two-dimensional (2D) perturbations (Shukla & Alam, J. Fluid Mech., vol. 672, 2011b, pp. 147–195). The differential equations for modal amplitudes (the fundamental mode, the mean-flow distortion, the second harmonic and the distortion of the fundamental mode), up to cubic-order in perturbation amplitude, are solved using a spectral-based numerical technique, yielding an estimate of the first Landau coefficient that accounts for the leading-order nonlinear effect on finite-amplitude perturbations. In the near-critical regime of flows, we found evidence of mean-flow resonance, characterized by the divergence of the first Landau coefficient, that occurs due to the interaction/resonance between a linear instability mode and a mean-flow mode. The nonlinear solutions are found to appear via both pitchfork and Hopf bifurcations from the underlying linear instability modes, leading to supercritical nonlinear states of stationary and travelling wave solutions. The subcritical travelling wave solutions have also been uncovered in the linearly stable regimes of flow. It is shown that multiple nonlinear states of both stationary and travelling waves can coexist for a given parameter combination of mean density and Couette gap. The 3D nonlinear solutions persist for a range of spanwise wavenumbers up to ${k}_{z} = O(1)$ that originate from 2D instabilities which occur beyond a moderate value of the mean density. For purely 3D instabilities in dilute flows (having no analogue in 2D flows), the supercritical finite-amplitude solutions persist for a much larger range of spanwise wavenumber up to ${k}_{z} = O(10)$. For all instabilities, the vortical motion on the cross-stream plane has been characterized in terms of the fixed/critical points of the underlying flow field: saddles, nodes (sources and sinks) and vortices have been identified. While the cross-stream velocity field for supercritical solutions in dilute flows contains nodes and saddles, the subcritical solutions are dominated by large-scale vortices in the background of saddle-node-type motions. The latter type of flow pattern also persists at moderate densities in the form of supercritical nonlinear solutions that originate from the dominant 2D instability modes for which the vortex appears to be driven by two nearby saddles. The location of this vortex is found to be correlated with the local maxima of the streamwise vorticity.

Curie point depth from spectral analysis of aeromagnetic data from Cerro Prieto geothermal area, Baja California, México

Using aeromagnetic data acquired in the area from the Cerro Prieto geothermal field, we estimated the depth to the Curie point isotherm, interpreted as the base of the magnetic sources, following statistical spectral-based techniques. According to our results the Curie point isotherm is located at a depths ranging from 14 to 17 km. Our result is somewhat deeper than that obtained previously based only in 2-D and 3-D forward modeling of previous low-quality data. However, our results are supported by independent information comprising geothermal gradients, seismicity distribution in the crust, and gravity determined crustal thickness. Our results imply a high thermal gradient (ranging between 33 and 38 °C/km) and high heat flow (of about 100 mW/m 2) for the study area. The thermal regime for the area is inferred to be similar to that from the Salton trough.

Significant EEG Features Involved in Mathematical Reasoning: Evidence from Wavelet Analysis

Using electroencephalographic (EEG) signals and a novel methodology based on wavelet measures in the time-scale domain, we evaluated cortex reactions during mathematical thinking. Our purpose was to extract more precise information from the cortex reactions during this cognitive task. Initially, the brain areas (lobes) of significant activation during the task are extracted using time-averaged wavelet power spectrum estimation. Then, a refinement step makes use of statistical significance-based criteria for comparing wavelet power spectra between the task and the rest condition. EEG signals are recorded from 15 young normal volunteers using 30 scalp electrodes as participants performed one difficult arithmetic task and the results are compared with a rest situation. The results are in accordance with similar previous studies, showing activations of frontal and central regions. Compared with the alternative spectral-based techniques, the method we propose achieves higher task discrimination on the same dataset and provides additional detail-signal information to evaluate cortical reactivity during local cortical activation.

Modeling of friction using spectral analysis

Usually friction is considered to be a velocity-dependent phenomenon. As a result, most of the available friction models describe friction only as a function of velocity. However, there is also positional dependency which is routinely ignored. Experiments reported in this paper confirm previously published findings of others that friction is position dependent, and that this dependency should not be ignored. Therefore, it is necessary to develop models that describe friction as a function of both position and velocity. Furthermore, available models of nonlinear friction are usually developed by investigating simple cases, such as sliding of two flat surfaces or rolling of a perfectly rounded body on a flat surface. These models do not accurately describe friction generated by mechanisms that have numerous sliding and rolling parts. In this paper, a new spectral-based modeling technique capable of describing nonlinear friction as a function of position and velocity is proposed. This technique identifies the specific contribution to the overall friction of every moving part of the mechanism.

Parallel adaptive rooting algorithm for general frequency estimation and direction finding

An adaptive rooting algorithm, is proposed for tracking the nonstationary roots of null spectra for data from sensor arrays, or time series. Without the assumption of uniformity for linear arrays, this adaptive-rooting algorithm associated with the zero-extraction technique is further extended to construct a parallel-effective-rooting processor. The adaptive algorithm tracks all the principal roots and reduces the computational complexity from order N3 to order N2. Simulations show better resolution performance of the suggested algorithm than that of the usual of spectral-based high-resolution techniques such as minimum-norm or MUSIC, for both uniform and nonuniform cases. Simulations also reveal that its convergence speed competes with those of recent fast adaptive eigen-structure algorithms for tracking the nonstationary parameters.