Application of sparse S transform network with knowledge distillation in seismic attenuation delineation

Time-frequency analysis is a successfully used tool for analyzing the local features of seismic data.However,it suffers from several inevitable limitations,such as the restricted time-frequency resolution,the difficulty in selecting parameters,and the low computational efficiency.Inspired by deep learning,we suggest a deep learning-based workflow for seismic time-frequency analysis.The sparse S transform network(SSTNet)is first built to map the relationship between synthetic traces and sparse S transform spectra,which can be easily pre-trained by using synthetic traces and training labels.Next,we introduce knowledge distillation(KD)based transfer learning to re-train SSTNet by using a field data set without training labels,which is named the sparse S transform network with knowledge distillation(KD-SSTNet).In this way,we can effectively calculate the sparse time-frequency spectra of field data and avoid the use of field training labels.To test the availability of the suggested KD-SSTNet,we apply it to field data to estimate seismic attenuation for reservoir characterization and make detailed comparisons with the traditional time-frequency analysis methods.

Automatic Delineation of Lung Parenchyma Based on Multilevel Thresholding and Gaussian Mixture Modelling

Delineation of the lung parenchyma in the thoracic Computed Tomography(CT)is an important processing step for most of the pulmonary image analysis such as lung volume extraction,lung nodule detection and pulmonary vessel segmentation.An automatic method for accurate delineation of lung parenchyma in thoracic Computed Tomography images is presented in this paper.The proposed method involves a segmentation phase followed by a lung boundary correction technique.The tissues in the thoracic Computed Tomography can be represented by a number of Gaussians.We propose a histogram utilized Adaptive Multilevel Thresholding(AMT)for estimating the total number of Gaussians and their initial parameters.The parameters of Gaussian components are updated by Expectation Maximization(EM)algorithm.The segmented lung parenchyma from the Gaussian Mixture model(GMM)undergoes an Adaptive Morphological Filtering(AMF)to reduce the boundary errors.The proposed method has been tested on 70 diseased and 119 normal lung images from 28 cases obtained from Lung Image Database Consortium(LIDC).The performance of the proposed system has been validated.

A Method for Alpine Wetland Delineation and Features of Border: Zoigê Plateau, China

Accurate wetland delineation is the basis of wetland definition and mapping, and is of great importance for wetland management and research. The Zoige Plateau on the Qinghai-Tibet Plateau was used as a research site for research on alpine wetland delineation. Several studies have analyzed the spatiotemporal pattern and dynamics of these alpine wetlands, but none have addressed the issues of wetland boundaries. The objective of this work was to discriminate the upper boundaries of alpine wetlands by coupling ecological methods and satellite observations. The combination of Landsat 8 images and supervised classification was an effective method for rapid identification of alpine wetlands in the Zoig6 Plateau. Wet meadow was relatively stable compared with hydric soils and wetland hydrology and could be used as a primary indicator for discriminating the upper boundaries of alpine wetlands. A slope of less than 4.5° could be used as the threshold value for wetland delineation. The normalized difference vegetation index （NDVI） in 434 field sites showed that a threshold value of 0.3 could distinguish grasslands from emergent marsh and wet meadow in September. The median normalized difference water index （NDWI） of emergent marsh remained more stable than that of wet meadow and grasslands during the period from September until July of the following year. The index of mean density in wet meadow zones was higher than the emergent and upland zones. Over twice the number of species occurred in the wet meadow zone compared with the emergent zone, and close to the value of upland zone. Alpine wetlands in the three reserves in 2014 covered 1175.19 kin2 with a classification accuracy of 75.6%. The combination of ecological methods and remote sensing technology will play an important role in wetland delineation at medium and small scales. The correct differentiation between wet meadow and grasslands is the key to improving the accuracy of future wetland delineation.

Improving the accuracy and consistency of clinical target volume delineation for rectal cancer by an education program

BACKGROUND Accurate target volume delineation is the premise for the implementation of precise radiotherapy.Inadequate target volume delineation may diminish tumor control or increase toxicity.Although several clinical target volume(CTV)delineation guidelines for rectal cancer have been published in recent years,significant interobserver variation(IOV)in CTV delineation still exists among radiation oncologists.However,proper education may serve as a bridge that connects complex guidelines with clinical practice.AIM To examine whether an education program could improve the accuracy and consistency of preoperative radiotherapy CTV delineation for rectal cancer.METHODS The study consisted of a baseline target volume delineation,a 150-min education intervention,and a follow-up evaluation.A 42-year-old man diagnosed with stage IIIC(T3N2bM0)rectal adenocarcinoma was selected for target volume delineation.CTVs obtained before and after the program were compared.Dice similarity coefficient(DSC),inclusiveness index(IncI),conformal index(CI),and relative volume difference[ΔV(%)]were analyzed to quantitatively evaluate the disparities between the participants’delineation and the standard CTV.Maximum volume ratio(MVR)and coefficient of variation(CV)were calculated to assess the IOV.Qualitative analysis included four common controversies in CTV delineation concerning the upper boundary of the target volume,external iliac area,groin area,and ischiorectal fossa.RESULTS Of the 18 radiation oncologists from 10 provinces in China,13 completed two sets of CTVs.In quantitative analysis,the average CTV volume decreased from 809.82 cm3 to 705.21 cm3(P=0.001)after the education program.Regarding the indices for geometric comparison,the mean DSC,IncI,and CI increased significantly,whileΔV(%)decreased remarkably,indicating improved agreement between participants’delineation and the standard CTV.Moreover,an 11.80%reduction in MVR and 18.19%reduction in CV were noted,demonstrating a smaller IOV in delineation after the education program.Regarding qualitative analysis,the greatest variations in baseline were observed at the external iliac area and ischiorectal fossa;61.54%(8/13)and 53.85%(7/13)of the participants unnecessarily delineated the external iliac area and the ischiorectal fossa,respectively.However,the education program reduced these variations.CONCLUSION Wide variations in CTV delineation for rectal cancer are present among radiation oncologists in China's Mainland.A well-structured education program could improve delineation accuracy and reduce IOVs.

Spatiotemporal Modeling of Monthly Precipitation in the Upper Shiyang River Watershed in West Central Gansu,Northwest China

This paper outlines a methodology to estimate monthly precipitation surfaces at 1-kin resolution for the Upper Shiyang River watershed （USRW） in northwest China. Generation of precipitation maps is based on the application of a four-variable genetic algorithm （GA） trained on 10 years of weather and ancillary data, i.e., surface air temperature, relative humidity, Digital Elevation Model-derived estimates of elevation, and time of year collected at 29 weather stations in west-central Gansu and northern Qinghai province. An observed-to-GA predicted data comparison of 10 years of precipitation collected at the 29 weather stations showed that about 84% of the variability in observed values could be explained by the trained GA, including variability in two independent datasets. Point-comparisons of observed and modeled precipitation along an elevation-rainfall gradient demonstrated near-similar spatiotemporal patterns. A precipitation surface for USRW for July, 2005, was developed with the trained GA and input surfaces of surface air temperature and relative humidity generated from Moderate Resolution Imaging Spectroradiometer sensor （MODIS） products of land surface temperature. Spatial tendencies in predicted maximum and minimum values of surface air temperature, relative humidity, and precipitation within a 2-kin radius circle around selected weather stations were in close agreement with the values measured at the weather stations.

Automated tree detection and crown delineation using airborne laser scanner data in heterogeneous East-Central Europe forest with different species mix

Identifying tree locations is a basic step in the derivation of other tree parameters using remote sensing techniques, particularly when using airborne laser scanning. There are several techniques for identifying tree positions. In this paper, we present a raster-based method for determining tree position and delineating crown coverage. We collected data from nine research plots that supported different mixes of species. We applied a raster-based method to raster layers with six different spatial resolutions and used terrestrial measurement data as reference data. Tree identification at a spatial resolution of 1.5 m was demonstrated to be the most accurate, with an average identification ratio (IR) of 95% and average detection ratio of 68% being observed. At a higher spatial resolution of 0.5 m, IR was overestimated by more than 600%. At a lower spatial resolution of 3 m, IR was underestimated at less than 44% of terrestrial measurements. The inventory process was timed to enable evaluation of the time efficiency of automatic methods.

Positron emission tomography and magnetic resonance imaging combined with computed tomography in tumor volume delineation: A case report

BACKGROUND Accurate delineation of the target area for patients with hypopharyngeal cancer is the key to achieving an ideal radiotherapy effect.Since computed tomography(CT)alone can no longer meet the treatment needs,fusing CT images with magnetic resonance imaging(MRI)or positron emission tomography(PET)images can overcome the disadvantages of CT.Herein,we present a clinical case of hypopharyngeal cancer to delineate the tumor volume using combined MRI-CT and PET-CT fusion images to examine if they could accurately cover the tumor volume.CASE SUMMARY A 67-year-old male patient with hypopharyngeal carcinoma could not tolerate chemotherapy and surgery due to complicated health issues such as diabetic nephropathy and other underlying diseases.After multidisciplinary consultations,clinicians eventually agreed to undergo radiotherapy to control the progression of his tumor.He was examined by CT,MRI,and 18-fluorodeoxyglucose-PET for treatment planning,and CT images were fused with PET and MRI images while delineating tumor volume.CONCLUSION The image fusion of MRI-CT and PET-CT has both advantages and disadvantages.Compared with CT images alone,the combination of MRI-CT and PET-CT fusion images can precisely cover the gross tumor volume in hypopharyngeal carcinoma and avoid overestimation or incomplete coverage of tumor volume.

Improving Curve Number Runoff Estimates Using Dual Hydrologic Soil Classification and Potential Contributing Source Areas Delineation Methods

Runoff models such as the Curve Number (CN) model are dependent upon land use and soil type within the watershed or contributing area. In regions with internal drainage (e.g. wetlands) watershed delineation methods that fill sinks can result in inaccurate contributing areas and estimations of runoff from models such as the CN model. Two methods to account for this inaccuracy have been 1) to adjust the initial abstraction value within the CN model;or 2) to improve the watershed delineation in order to better account for internal drainage. We used a combined approach of examining the watershed delineation, and refining the CN model by the incorporating of dual hydrologic soil classifications. For eighteen watersheds within Wisconsin, we compared the CN model results of three watershed delineation methods to USGS gaged values. We found that for large precipitation events (>100 mm) the CN model estimations are closer to observed values for watershed delineations that identify the directly connected watershed and use the undrained hydrologic soil classification.

Delineation of Haigou gold metallogenic province and its metallogenic prospection

Based on the accumulated data for the gold deposits in the central Jilin Province in recent years and our understanding of the gold metallogenic province,the Haigou gold metallogenic province is delineated and the denudation degree of gold deposits in this province is discussed. The potential and the ore-searching direction of the province are also considered. The Haigou gold metallogenic province occurs as an independent province with low denudation degree and high ore-producing potential. Regional fault belts and small basic intrusions are two ore-constrains and could serve as the ore-searching indictors in the province.

A Preliminary Study of Automatic Delineation of Eyes on CT Images Using Ant Colony Optimization

Eyes are important organs-at-risk （OARs） that should be protected during the radiation treatment of those head tumors. Correct delineation of the eyes on CT images is one of important issues for treatment planning to protect the eyes as much as possible. In this paper, we propose a new method, named ant colony optimization （ACO）, to delineate the eyes automatically. In the proposed algorithm, each ant tries to find a closed path, and some pheromone is deposited on the visited path when the ant fmds a path. After all ants fmish a circle, the best ant will lay some pheromone to enforce the best path. The proposed algorithm is verified on several CT images, and the preliminary results demonstrate the feasibility of ACO for the delineation problem.

LIMITATIONS OF CANCER MARGIN DELINEATION BY MEANS OF AUTOFLUORESCENCE IMAGING UNDER CONDITIONS OF LASER SURGERY

Limitations of cancer margin delineation and surgical guidance by means of autofluorescence imaging under conditions of laser ablation were investigated and preliminary results are presented.PinPoint^(TM)(Novadaq Technologies Inc.,Canada)was used to capture digital images and Er:YAG laser(2.94μm,Glissando,WaveLight^(TM),Germany)was exploited to cause laser ablation on both normal and cancer sites of the specimen.It was shown that changes of the autofluorescence image after ablation extend beyond the actual sizes of the ablation loci.The tumor tissue after the laser ablation starts to emit fluorescent light within the green wavelength band(490-550nm)similar to normal tissue stating that the current technology of in-process tissue classification fails.However,when the autofluorescence was collected in the red range(600-750nm),then the abnormal/normal contrast was reduced,but still present even after the laser ablation.The present study highlights the importance of finding a proper technology for surgical navigation of cancer removal under conditions of high power effects in biological tissues.

Application of digital elevation model in delineating drainage networks

A practical method to extract drainage network from DEM (digital elevation model) is introduced. DEM pretreatment includes depression and flat areas treatment. The flow direction of each grid cell in DEM is calculated according to the 8-direction pour point model, and then the flow accumulation grid from the flow direction grid. With the flow accumulation grid, streams are defined according to the given threshold value of flow accumulation. Taking Gufo River watershed as an example, the extraction of drainage network was done from DEM. The results are basically consistent with the digitized drainage network from the relief maps.

Complex Target Volume Delineation and Treatment Planning in Radiotherapy for Malignant Pleural Mesothelioma (MPM)

Introduction: Radiotherapy alone or combined with surgery and/or chemotherapy is being investigated in the treatment of malignant pleural mesothelioma (MPM). This study aimed to simulate a Volumetric Modulated Arc Therapy (VMAT) treatment of a patient with MPM. Materials and Methods: CT images from a patient with intact lungs were imported via DICOM into the Pinnacle3 treatment planning (TP) system (TPS) and used as a model for MPM to delineate organs at risk (OAR) and both clinical and planning target volumes (CTV and PTV) with a margin of 5 mm. Elekta Synergy with 6 MV photons and 80 leafs MLCi2 was employed. VMAT plans were generated using two coplanar arcs with gantry rotation angles of 178° - 182°, the collimator angles of each arc were set to 90°, Octavius® 4D 729 was employed for quality assurance while the calculated and measured doses were compared using VeriSoft. Results: A TP was achieved. The Gamma volume analysis with criteria of 3 mm distance to agreement and 3% dose difference yielded the gamma passing rate = 99.9%. The reference isodose was 42.75 Gy with the coverage constraints for the PTV D95 and V95 = 95.0% of 45 Gy. The remaining dosimetric parameters met the recommendations from the clinically acceptable guidelines for the radiotherapy of MPM. Conclusion: Using well-defined TV and VMAT, a consistent TP compared to similar ones from publications was achieved. We obtained a high agreement between the 3D dose reconstructed and the dose calculated.

Structural Connectivity Enhanced Anisotropic 3D Network for Brain Midline Delineation

Brain midline delineation can facilitate the clinical evaluation of brain midline shift,which has a pivotal role in the diagnosis and prognosis of various brain pathology.However,there are still challenges for brain midline delineation:1)the largely deformed midline is hard to localize if mixed with severe cerebral hemorrhage;2)the predicted midlines of recent methods are not smooth and continuous which violates the structural priority.To overcome these challenges,we propose an anisotropic three dimensional(3D)network with context-aware refinement(A3D-CAR)for brain midline modeling.The proposed network fuses 3D context from different two dimensional(2D)slices through asymmetric context fusion.To exploit the elongated structure of the midline,an anisotropic block is designed to balance the difference between the adjacent pixels in the horizontal and vertical directions.For maintaining the structural priority of a brain midline,we present a novel 3D connectivity regular loss(3D CRL)to penalize the disconnectivity between nearby coordinates.Extensive experiments on the CQ dataset and one in-house dataset show that the proposed method outperforms three state-of-the-art methods on four evaluation metrics without excessive computational burden.

New delineation of two favorable zones for gas hydrate in southem Qinghai and northern Tibet,China

1.ObjectivesSouthern Qinghai-northem Tibet permafrost region is a place having the most widespread and most developed permafrost in China with good mineralization conditions and prospecting potentials for gas hydrate (Zhu YH et al.,2011). In 2011,China Geological Survey initiated a special national program entitled "Gas hydrate resource exploration and trial mining",which signaled a prelude to a comprehensive gas hydrate survey in southern Qinghai-northern Tibet permafrost region.So far,appreciable progress has been made in the geological,geophysical,geochemical and drilling survey across a number of key blocks.This paper is intended to examine the fundamental conditions for gas hydrate mineralization based on previous findings,delineate favorable zones for gas hydrate mineralization,and to make contributions to the onshore gas hydrate resource exploration in China.

The Current Status of the Delineation and Determination of the Targets and the Radiation Protocols for Nasopharyngeal Cancer with Intensity-Modulated Radiotherapy

IMRT has increased the local-regional control and decreased the complications from treating nasopharyngeal cancer （NPC）. Therefore studying IMRT is important. CT and MRI are complementary, and their joint use is currently considered to be the optimal modality to delineate the extent of the primary spread of NPC. The key problem in delineation of the neck nodes is how to translate anatomic node regions into the CT boundaries. The consensus guideline which narrowed the gap among different cancer centers is recommended in delineating the boundary of the cervical lymph node regions. The definition of the NPC GTV is clear and almost the same among the main cancer centers in their IMRT planning protocols. The suggested biological dose to the GTV is close to or more than 80 Gy; the main differences are the definitions of the CTVs and their schemes for the prescribed dose, and also the dosage to the high cervical region is different among those centers. According to their long-term follow-up results, it is suggested that, besides adding 5-10 mm margins to the primary lesions, the immediate high-risk structures （including the entire nasopharyngeal cavity, retropharyngeal space, clivus, base of the skull, pterygoid plates and muscles, parapharyngeal space, the sphenoid and partial ethmoid sinuses, the posterior third of the maxillary sinuses and the nasal cavity） should also be included with a prescription of more than 60 Gy, and the bilateral Ib, II and Va node levels should be ranked as high-risk regions and differentially prescribed for treatment with no less than 60 Gy.

Delineation of Mesoscale Features of Ocean on Satellite IR Image

An ICSED (Improved Cluster Shade Edge-Detection) algorithm and a series of post-processing technique are discussed for automatic delineation of mesoscale structure of the ocean on digital IR images. The popular derivative-based edge operators are shown to be too sensitive to edge fine-structure and to weak gradients. The new edge-detection algorithm is ICSED (Improved Cluster Shade Edge-detection) method and it is found to be an excel lent edge detector that exhibits the characteristic of fine-structure rejection while retaining edge sharpness. This char acteristic is highly desirable for analyzing oceanographic satellite images. A sorting technique for separating clouds or land well from ocean at both day and night is described in order to obtain high quality mesoscale features on the IR image This procedure is evaluated on an AVHRR (Advanced Very High Resolution Radiometer) image with Kuroshio. Results and analyses show that the mesoscale features can be well identified by using ICSED algorithm.

Review of the Water Footprint Project within Geographically Delineated Area

During the last decade, there has been an intensive research activity concerning the concept of the Water Footprint （WF） approach, which was firstly introduced by Arjen Hoekstra in 2002. WF is an indicator of direct and indirect freshwater use of a consumer or producer that takes into account water consumption in every step （intermediate and final） along the production chain and services. The concept can be implemented in various levels such as products, consumers, producers, nations and river basins etc.. The water footprint within a geographically delineated area equals the sum of the process water footprints of all processes taking place in the area. The aim of current research is a review of the most important WF studies, with a special focus on applications within regional, basin and administrative unit level. National and global scales are not included in the current paper. The article presents the most widespread methodologies and approaches that attempt to evaluate water footprints of specific defined areas and highlights their recent advances as well as shortcomings in the constantly evolving research efforts.

The Dilemma of Target Delineation with PET/CT in Radiotherapy Planning for Malignant Tumors

Currently there are many unanswered questions concerning contouring a target with PET/CT in radiotherapy planning. Who should contour the PET volume-the radiation oncologist or the nuclear medicine physician？ Which factors will contribute to the dual-observer variability between them？ What should be taken as the optimal SUV threshold to demarcate a malignant tumor from the normal tissue？ When the PET volume does not coincide with the local area CT findings, which portion should be contoured as the target？ If a reginal lymph node,draining area or a remote region is shown to be PET positive but CT negative, or PET negative but CT positive, how is the target identified and selected？ Further studies concerning the relationship between PET/CT and the cancerous tissue are needed. The long-term clinical results showing an increased therapeutic ratio will finally verify the applicability of guidelines to contour the target with PET/CT in radiotherapy planning.