Sub-millimeter Level Research Articles

Dynamic displacements from high-rate GNSS: Error modeling and vibration detection

Nowadays, GNSS (Global Navigation Satellite System) receivers can measure with sampling rates up to 100 Hz, and it is possible to provide displacements at a level of precision of a few millimeters. This opened an interesting field of applications, e.g. for disciplines that monitor dynamic movements, like seismology or structural health monitoring. However, the detection of smallest displacements possible requires a thorough analysis of both error types, systematic and stochastic. In view of this, this paper deals with the detection of small vibrations with amplitudes down to the sub-millimeter level, and the error modelling necessary to achieve this. The data under investigation are time series of kinematic GNSS displacement measurements. We will use parametric time series models of the ARIMA (Autoregressive Integrated Moving Average) type to characterize the GNSS errors and to do a pre-filtering based on a calibrated model. With non-parametric spectral estimation methods – namely periodograms – we can then test for the presence of harmonic processes (sinusoids) in the data. Once a sinusoid is detected, its amplitude and phase are estimated by means of harmonic regression. Subsequently, the sinusoid can be removed, and the procedure can be repeated. This algorithm is tested with data collected for a GNSS shake table experiment; it is shown that it is possible resolve and detect vibrations down to a level of tens of microns with few minutes of data being analyzed. We also highlight some theoretical considerations on minimum detectable amplitudes of vibrations for measurements of different quality and record lengths.

Monitoring and Analysis of Dynamic Characteristics of Super High-rise Buildings using GB-RAR: A Case Study of the WGC under Construction, China

Accurate dynamic characteristics of super high-rise buildings serve as a guide in their construction and operation. Ground-based real aperture radar (GB-RAR) techniques have been applied in monitoring and analyzing the dynamic characteristics of different buildings, but only few studies have utilized them to derive the dynamic characteristics of super high-rise buildings, especially those higher than 400 m and under construction. In this study, we proposed a set of technical methods for monitoring and analyzing the dynamic characteristics of super high-rise buildings based on GB-RAR and wavelet analysis. A case study was conducted on the monitoring and analysis of the dynamic characteristics of the Wuhan Greenland Center (WGC) under construction (5–7 July 2017) with a 636 m design height. Displacement time series was accurately derived through GB-RAR and wavelet analysis, and the accuracy reached the submillimeter level. The maximum horizontal displacement amplitudes at the top of the building in the north–south and east–west directions were 18.84 and 15.94 mm, respectively. The roof displacement trajectory of the WGC was clearly identified. A certain negative correlation between the temperature and displacement changes at the roof of the building was identified. Study results demonstrate that the proposed method is effective for the dynamic monitoring and analysis of super high-rise buildings with noninvasive and nondestructive characteristics.

Multispectral tissue mapping - a novel concept for the optical evaluation of liver malignancy

Introduction: Malignancy alters the optical absorption behaviour of liver tissue. This process can be evaluated using multi spectral analysis, a method that has been proposed as an intraoperative diagnostic tool. Existing systems either require invasive probe insertion or have limited spectral sensitivity which reduces their diagnostic ability. This work describes the development and evaluation of a novel, imaging method called multispectral tissue mapping, which addresses these issues. Methods: The imaging system (spectral range 1100-1700nm) consists of a tuneable excitation light source and a near-infrared camera. Following ethics approval, absorption spectra from resected human liver specimen were acquired ex-vivo. Twenty repeat acquisitions of malignant and non-malignant tissues from the same patient were averaged to reduce variability and subsequently compared. Generalised linear mixed modelling was employed for statistical analysis of serial changes in pixel based absorption (pixel area≈0.5mm2). Subsequently absorption data was transformed into augmented colour images with the aim of identifying cancer tissue. Results: Multiple spectra from liver specimens with colorectal cancer liver metastasis (CRLM) (three patients, n=505.141 pixel values) and hepatocellular carcinoma (HCC) (two patients, n=674.477 pixel values) were examined. Malignant tissue had a heterogeneous appearance on multispectral imaging compared to non-malignant tissue. Whereas absorption was increased in CRLM (maxima 1410nm; OR 0.005, CI 0.004-0.007) it was reduced in HCC (minimum 1440nm; OR 5.4, CI 3.9-7.6). Data was randomised into two groups and subsequently a predictive algorithm was built and tested. On ROC analysis predictive accuracy was 0.6 (area under curve) for HCC and 0.7 for CRLM, respectively. Augmented colour processing empirically enhanced image contrast between malignant and non-malignant tissue. Conclusion: Multispectral tissue mapping enables identification of liver malignancy on a sub-millimetre level. Augmented colour representation may enhance image contrast between malignant and normal liver tissues. Studies on larger datasets are required to improve the accuracy of the predictive algorithm.

Ocean Tide Loading Displacement Parameters Estimated From GNSS-Derived Coordinate Time Series Considering the Effect of Mass Loading in Hong Kong

In order to analyze the influence of mass loading [atmospheric loading (ATML), hydrological loading (HYDL), and nontidal ocean loading (NTOL)] on estimating ocean tide loading displacement (OTLD) parameters by GNSS, the kinematic precise point positioning technique is used to process the GNSS data in Hong Kong to obtain coordinate time series. The 3-D OTLD parameters of eight principal constituents (M2, S2, N2, K2, K1, O1, P1, and Q1) are estimated using harmonic analysis method from the coordinate time series with/without mass loading correction and random noise reduction. The GNSS-derived OTLD parameters are compared with the FES2014 and NAO.99b. The results show that GNSS/model agreement is generally at the submillimeter level (excluding S2 and K1). After mass loading is corrected, the residual between OTLD parameter by GNSS and model of M2, S2, N2, P1, and Q1 have been reduced more than 20%, while the accuracy improvement of other tidal constituents is not obvious. The influence of different mass loading on estimation of OTLD parameters is analyzed. Compared with ocean tide model, root mean square error misfit decreased by 7.52%, 7.76%, and 7.58% in north (N), east (E), and up (U) with ATML correction, respectively. The accuracy is improved little in vertical direction and decreases slightly in horizontal direction with NTOL correction. The HYDL correction has little effect on OTLD estimation. After the random noise reduction, the accuracy of the estimation of OTLD parameters has been significantly improved, and the improvement of the accuracy of the semidiurnal constituents is significantly greater than those of the diurnal constituents.

Absolute field calibration for multi-GNSS receiver antennas at ETH Zurich

ETH Zurich developed an absolute GNSS antenna calibration system based on measurements taken in the field. An industrial robot is used to rotate and tilt the antenna to be calibrated. This procedure ensures good coverage of the antenna hemisphere and reduces systematic errors. The calibration system at ETH Zurich is validated by a direct comparison of the obtained calibrations with calibrations from the anechoic chamber method (University of Bonn) and from another absolute field calibration method (Geo++® GmbH). Calibrations by ETH Zurich agree on the sub-millimeter level with both reference calibrations. A second validation was conducted using real measurements on short baselines. Data were acquired on four stations in direct vicinity and processed using different phase center correction models. The experiment shows that individual corrections of ETH Zurich reduce the residuals in the coordinate domain when compared to type-mean calibrations of the International GNSS Service (IGS). However, residual biases between GPS and Galileo coordinates remain. These biases are efficiently reduced when using the new type-mean calibrations from the IGS that include calibration values for all GNSS, including Galileo. The ETH Zurich calibration system is proven to deliver meaningful calibrations that agree with other calibrations on the millimeter level in the azimuth and elevation domain. The field validation shows evidence that the consistency of the Galileo and GPS calibration should be further enhanced by performing a combined GPS and Galileo analysis, which is not yet implemented.

Design of multipath error correction algorithm of coarse and fine sparse decomposition based on compressed sensing in time-of-flight cameras

ABSTRACTA single pixel in time of flight cameras receives multiple reflected light from different scene points, resulting in erroneous depth information. In this paper, based on the proposed sparse decomposition, a coarse and fine sparse decomposition based on compressed sensing is applied to multipath separation. The applied method uses a linear combination of multiple frequency signals to modulate the source. The measured vector obtained through finite random measurements is subjected to two sparse decompositions – rough separation and detailed positioning, and finally the minimum direct path depth is accurately recovered. Under the premise of the same number of measurements, calculation amount, and storage space, the accuracy of the coarse and fine sparse decomposition based on compressed sensing is improved by nearly an order of magnitude compared to the sparse decomposition without compressed sensing. Moreover, our method can basically achieve multi-path separation accuracy to the sub-millimeter level.

DEVELOPMENT OF A HIGH-SPEED VIDEOGRAMMETRIC MEASUREMENT SYSTEM WITH APPLICATION IN LARGE-SCALE SHAKING TABLE TEST

Abstract. The high-speed videogrammetric measurement system, which provides a convenient way to capture three-dimensional (3D) dynamic response of moving objects, has been widely used in various applications due to its remarkable advantages including non-contact, flexibility and high precision. This paper presents a distributed high-speed videogrammetric measurement system suitable for monitoring of large-scale structures. The overall framework consists of hardware and software two parts, namely observation network construction and data processing. The core component of the observation network is high-speed cameras to provide multiview image sequences. The data processing part automatically obtains the 3D structural deformations of the key points from the captured image sequences. A distributed parallel processing framework is adopted to speed up the image sequence processing. An empirical experiment was conducted to measure the dynamics of a double-tube five-layer building structure on the shaking table using the presented videogrammetric measurement system. Compared with the high-accuracy total station measurement, the presented system can achieve a sub-millimeter level of coordinates discrepancy. The 3D deformation results demonstrate the potential of the non-contact high-speed videogrammetric measurement system in dynamic monitoring of large-scale shake table tests.

A supervised network for fast image-guided radiotherapy (IGRT) registration.

3D/3D image registration in IGRT, which aligns planning Computed Tomography (CT) image set with on board Cone Beam CT (CBCT) image set in a short time with high accuracy, is still a challenge due to its high computational cost and complex anatomical structure of medical image. In order to overcome these difficulties, a new method is proposed which contains a coarse registration and a fine registration. For the coarse registration, a supervised regression convolutional neural networks (CNNs) is used to optimize the spatial variation by minimizing the loss when combine the CT images with the CBCT images. For the fine registration, intensity-based image registration is used to calculate the accurate spatial difference of the input image pairs. A coarse registration can get a rough result with a wide capture range in less than 0.5s. Sequentially a fine registration can get accurate results in a reasonable short time. RSD-111T chest phantom was used to test our new method. The set-up error was calculated in less than 10s in time scale, and was reduced to sub-millimeter level in spatial scale. The average residual errors in translation and rotation are within ±0.5mm and ± 0.2°.

TLS FOR DETECTING SMALL DAMAGES ON A BUILDING FAÇADE

Abstract. Weathering, aging, infiltration, solar radiation and several other factors cause the deterioration of buildings and infrastructures and hence the need for periodical maintenance and restoration. The need for maintenance has been traditionally determined based on visual inspections of qualified operators. Since this process is obviously time consuming and quite expensive, especially when the considered building is quite large and high, then a number of recent studies have been recently published proposing remote sensing tools in order to ease the monitoring process. Among the possible spatial data acquisition sensors, terrestrial laser scanning has been considered in several of the existing studies, mostly because of its high reliability, to cope with cracks and defect detection up to the millimeter level of resolution, which is the typical accuracy of the current generation of professional laser scanners. This paper considers the problem of detecting small defects on the façade of a University building. Similar to other previous studies, in this work defect detection is accomplished by considering distances with respect to a planar surface locally fitted on the building façade. Then, statistical filtering and machine learning tools have been implemented in order to cope with damage detection of the brick surfaces at sub-millimeter level.

HIGH RESOLUTION 3D ACQUISITION AND MODELLING IN CULTURAL HERITAGE KNOWLEDGE AND RESTORATION PROJECTS: THE SURVEY OF THE FOUNTAIN OF NEPTUNE IN BOLOGNA

Abstract. In 2016, the Municipality of Bologna (Italy) has undertaken the restoration of one of the symbols of the entire city, the Fountain of Neptune, in evident state of degradation. The works have touched upon all the aspects of this complex object and the project has seen involved the Municipality and the University of Bologna, the Istituto Superiore per la Conservazione e il Restauro (ISCR) (Rome) and the Visual Computing Lab of the CNR – ISTI (Pisa), in a modern and highly multi-disciplinary approach.One of the key elements of the project was made up by the creation of an information system ad hoc developed to permit, in an innovative, efficient and user-friendly way, the collection, sharing, management and analysis of all the information and data related to diagnostics and restoration actions. The base of the information system is a very detailed 3D model of the monument, realized by means of the most modern techniques for objects 3D modelling (laser scanning, digital photogrammetry and 3D scanning) integrated together with the aim to obtain a photo-textured 3D model characterized by a sub-millimetre precision level in the geometric description and a high perceptive fidelity of colour reproduction.The surveying activities and data processing, performed by the DICAM Geomatics group of the University of Bologna (with the collaboration of the MCM Company of Rome), are described in the paper, with considerations on the problems encountered and the procedures and solutions adopted. The information system has been developed by CNR-ISTI.

Airborne Distributed POS Flexible Baseline Measurement Method Based on MCLS

The distributed position and orientation measurement system (DPOS) depends on transfer alignment to accurately measure multi-node motion information for multi-task imaging sensors. However, the transfer alignment accuracy is limited by the flexible baseline accuracy between nodes. A flexible baseline measurement method based on the multiple camera laser-rangefinder measurement system (MCLS) is proposed in this paper. The relative position and orientation information between the nodes are measured directly by using MCLS. Combined with the advantages of visual measurement and laser ranging, this method completes the non-contact measurement and overcomes the difficult problem of the poor precision of the flexible baseline caused by inaccurate mathematical modeling. In order to verify this method, a distributed flexible baseline simulation platform is built and static test experiments are carried out. It shows that this method can realize sub-millimeter level measurement. At the same time, it can provide a strong support for the application of DPOS.

Validation of Geometric and Dosimetric Accuracy of Edge Accelerator Gating With Electromagnetic Tracking: A Phantom Study

This work was to devise a comprehensive testing scheme to evaluate the geometric and dosimetric accuracy of the Edge accelerator gating with electromagnetic tracking (EMT) for its safety in clinical application. A CIRS thorax phantom was scanned with four-dimensional cone-beam CT (4D-CBCT) on an Edge accelerator while the simulated tumor was simultaneously tracked with an EMT system using Calypso. The geometric accuracy was validated by comparing the motion trajectories derived from Calypso and 4D-CBCT with the ground truth from motion control software. The two-dimensional and three-dimensional dynamic doses were measured with the Matrixx and ArcCHECK installed on a motion platform, both with and without EMT. For tumor motion with 5, 7.5 mm amplitudes, the average absolute differences of sample position between Calypso and the ground truth were 0.286±0.234 mm, 0.407±0.331 mm respectively. Dosimetric accuracy was validated with 3 mm/3% gamma criterion. The average gamma pass rates of 2D dynamic dose validation based on Matrixx were less than 46% without EMT, 97.3% using 2 mm gating limit, 96% using 3 mm gating limit and 93.4% using 5mm gating limit respectively. The mean 3D dynamic dose validation pass rates based on ArcCHECK were 65.9% without EMT, 96.2% using 3 mm gating limit, and 92.5% using 5 mm gating limit with EMT respectively. The geometric accuracy of the Calypso system in tracking the moving target area was stable at the submillimeter level. The dosimetric accuracy could be improved significantly with EMT using an appropriate gating limit.

Wide-Field Planar Magnetic Imaging Using Spins in Diamond

The nitrogen-vacancy color centers in diamond can be used as an excellent magnetic sensor. In this work, we present a wide-field planar magnetic imaging technique based on nitrogen-vacancy color centers. Through optically detected magnetic resonance (ODMR) which is readout from a scientific CMOS camera to extract resonance frequency, we experimentally reconstruct the magnetic field distribution of permanent magnets and particles of magnetite (Fe3O4). Moreover, by using this method we imagined the magnetic field distribution of one integrated cell heater widely used in temperature control with magnetic field suppression. And the experiment shows that this technique can be applied in engineering. The imaging field of this technology can reach the sub-millimeter level and the spatial resolution can reach the optical diffraction limit. The presented wide-field NV imaging technique offers wide-field imaging, high spatial resolution, and high magnetic sensitivity.

Determination of number and locations of ultrasound sensors needed to accurately track scapular motion in vivo

ABSTRACTUltrasonic imaging is a safe, established and non-invasive method to visualise the position of tissues within the body. Coupling existing skin mounted motion capture markers to ultrasound sensors could allow for precise measurement of bone positions and orientations, without the loss in accuracy due to relative displacement between skin and bone present in standard motion capture techniques. This is especially important in motions where skin-based markers alone are impractical or inaccurate, as for tracking scapular position when shoulder abduction is greater than 80 degrees. It is currently unclear how many ultrasound-coupled skin-mounted markers would be required to provide accurate scapular tracking, and where they should be placed. In this study, ultrasound-coupled skin marker quantities and positions were optimised using Monte Carlo and simulated annealing methods. Monte Carlo simulation was used to determine the average target registration error in simulated scapular plane abduction motions for groups ranging from 1 to 40 ultrasound-coupled sensors. After the range of ideal sensor quantities was identified, a simulated annealing algorithm was used to identify optimal positions for the sensors. Using these positions, a single skin-mounted, ultrasound-coupled sensor placed over the acromion was found to reduce simulated registration errors by 40.5% compared to regular skin-mounted markers. Groups consisting of three or more sensors reduced simulated registration error at maximum scaption to sub-millimeter levels.

Proposal for an accurate TMS-MRI co-registration process via 3D laser scanning

An important technical issue in transcranial magnetic stimulation (TMS) usage is how accurately the specific brain areas activated by TMS are assessed. However, in practice, electric field induced in TMS is dispersed and therefore actual estimation is still difficult. As a preliminary step, the projection line which is perpendicular to the TMS stimulation coil beneath the center of the coil must be accurately estimated into the brain. Therefore, we have developed a new TMS-MRI co-registration procedure that employs a 3D laser-scanner system that is very useful for general hand-manipulated TMS, and which easily estimates the TMS projection point onto the brain. The proposed system accurately captures the positional relationship between the TMS coil and anatomical images. The results of 3D image processing revealed that the registration error at each stage was kept within the submillimeter level. In addition, a motor evoked potential experiment examining the right finger motor area revealed that understandable responses were obtained when stimulation was targeted to the three different motor areas according to Penfield’s map. 3D laser scanning is a technique of substantial recent interest for anatomical co-registration. The proposed method demonstrated submillimeter level accuracy of TMS-MRI co-registration.

Impulse response timing differences in BOLD and CBV weighted fMRI

Recent advances in BOLD fMRI scan techniques have substantially improved spatial and temporal resolution, currently reaching to sub-millimeter and sub-second levels respectively. Unfortunately, there remain physiological barriers that prevent achieving this resolution in practice. BOLD contrast relies on the hemodynamic response to neuronal activity, whose associated cerebral blood oxygenation (CBO) changes may spread over several millimeters and last several seconds. Recent reports have suggested that significant improvements may be possible with cerebral blood volume (CBV)-weighted fMRI, which highlights the CBV changes rather than the BOLD changes associated with the hemodynamic response. Nevertheless, quantitative comparisons between CBV and BOLD are sparse, in particular regarding their temporal characteristics in human brain.To address this, we studied a cohort of subjects that received injection of ferumoxytol, an intravascular iron-oxide based contrast agent that introduces strong CBV contrast. An event-related visual stimulus paradigm was used to compare the impulse response (IR) for CBV and BOLD contrast, obtained with and without ferumoxytol, respectively. Experiments performed at 7 T (n = 5) at 1.2–1.5 mm spatial and 1 s temporal resolution showed that the onset time and time-to-peak of the CBV IR averaged 0.8 and 3.5 s respectively, both 0.6 s shorter than the BOLD IR. While significant, these improvements are relatively small and not expected to lead to practical advantages for the extraction of temporal information about neural activity.Nonlinearities in the observed IR were also compared and found to be similar between the CBV and BOLD, indicating that these are likely not caused by a ‘ceiling’ effect in the CBO response, but rather support a previously proposed model of vascular compliance, in which changes in vascular tone elicited by a preceding stimulus affect the IR.

Jun ware glazes: Chemistry, nanostructure and optical properties

A set of the Classic and official Jun ware shards found at various kilns at Yuzhou (Henan) are analyzed. The blue color and milky appearance of the glazes subject of intense debate as being due to either the glaze nanostructure or chemistry. The presence of submillimetre areas showing different color and opacity gives to the glazes a three dimensional cloudy sky appearance. Nanostructure and chemistry of the glazes, oxidation state of iron and nature of the copper nanoprecipitates are studied at a submillimetre level. Images of the glaze nanostructures are obtained using a Focus Ion Beam and Scanning Electron Microscope with nanometric resolution. The size and volume fraction of the nanostructures determined by image treatment are related to the chemical composition of the glazes. Differences between kiln productions are found and the origin of the color and opacity of the glazes discussed in terms of the chemistry and nanostructure.

Relative positioning of formation-flying spacecraft using single-receiver GPS carrier phase ambiguity fixing

In recent years, differential carrier phase-based relative positioning (or “baseline determination”) with precision at the millimeter and submillimeter levels has been demonstrated for the GRACE, TanDEM-X and Swarm missions in offline processing. Specific features of such missions have included the use of spacecraft of similar shapes placed in almost identical orbits as well as the use of consistent geodetic-class GPS receivers. These elements have proven to be advantageous for the computation of baseline solutions with such precision levels. Particularly, they have allowed to fully leverage the use of differential GPS techniques, including the estimation and use of carrier phase integer ambiguities. Similarly, the aforementioned spacecraft and orbit characteristics have made it possible to tightly constrain the relative dynamics of formations in the generation of reduced-dynamic solutions. Other than the former examples, prospective formation-flying mission proposals, such as SAOCOM-CS and PICOSAR, may comprise spacecraft with very different characteristics, including dissimilar GPS/GNSS receivers. Such cases may no longer provide favorable conditions for relative orbit determination strategies. As an alternative, absolute orbit solutions may be computed individually for each spacecraft and used for the generation of precise baseline products. This study aims at the assessment of the potential of single-receiver ambiguity fixing for the generation of precise baseline solutions. Results using flight data from the GRACE, TanDEM-X and Swarm missions exhibit baseline accuracy better than 5 mm (3D RMS) for a one-month test period in June 2016. As such, the presented solutions may be considered for prospective formation-flying remote sensing missions with baseline precision requirements at the subcentimeter level. Likewise, the method is considered of particular interest for future multi-spacecraft formations and swarms that require efficient determination of a large number of individual baselines.

Construction and comparison of 3D multi-source multi-band models for cultural heritage applications

Abstract 3D multi-band/multi-spectral reconstruction is a technique, which allows the use of images taken in different spectral bands for the reconstruction of spectrally-resolved 3D models of paintings. In this work, we propose a procedure for constructing and comparing 3D multi-band models obtained starting from a variety of input data (RGB and IR images, UV-induced fluorescence images, etc.) and/or acquired with different instruments. The procedure has been tested on two case studies, with the aim to highlight the critical issues associated to the specificities of the subject and of the environmental constrains: a 14th century panel painting signed by Barnaba from Modena, preserved at the Museum of San Matteo in Pisa, Italy and a 15th century panel painting preserved at the Regional Gallery of Palazzo Bellomo in Syracuse, Italy. RGB, IR and UV–VIS fluorescence models obtained with a conventional digital camera (RGB and UV) and a multisband camera (IR) were realized and compared, for revealing differences at sub-millimeter level.

A new strategy of stochastic modeling aiming at BDS hybrid constellation in precise relative positioning

A realistic stochastic model is the prerequisite in global navigation satellite system (GNSS) positioning applications. Considering that Beidou satellite navigation system (BDS) consists of geostationary earth orbit (GEO), inclined geosynchronous satellite orbit (IGSO) and medium earth orbit (MEO) satellites operating in different orbits with different heights, only one model being used in stochastic modeling is probably inadequate, which is the usual case for Global Positioning System (GPS). In this contribution, a new strategy of stochastic modeling for hybrid constellation BDS precise relative positioning is proposed, namely multiple elevation-dependent models (MEDM). In MEDM, parameters of three elevation-dependent models for GEO, IGSO, MEO satellites are estimated independently. After that, variances of phase and code observations for different orbits satellites can be estimated, then the stochastic model specific for BDS can be constructed in real time. The experimental results reveal that, when MEDM is compared with the sole elevation-dependent model and the carrier-to-noise power density ratios (C/N0)-dependent model, the maximum improvements in ambiguity resolution success rates (SR) in 5, 6, 7, 8 available satellites cases are about 0.2%, 13%, 3%, 1% for B1 frequency, and 1%, 6%, 2.5%, 0.2% for B2 frequency, respectively. The overall improvements in baseline solutions by using MEDM are in millimeter or submillimeter level. Hence, MEDM is more suitable for hybrid constellation of BDS in precise relative positioning applications.