Search

Literature on medical imaging segmentation claims that hybrid UNet models containing both Transformer and convolutional blocks perform better than purely convolutional UNet models. This recently touted success of hybrid Transformers warrants an investigation into which of its components contribute to its performance. Also, previous work has a limitation of analysis only at fixed dataset scales as well as unfair comparisons with other models where parameter counts are not equivalent. Here, we investigate the performance of a hybrid Transformer network i.e. the nnFormer for organ segmentation in prostate CT scans. We do this in context of replacing its various components and by constructing learning curves by plotting model performance at different dataset scales. To compare with literature, the first experiment replaces all the shifted-window(swin) Transformer blocks of the nnFormer with convolutions. Results show that the convolution prevails as the data scale increases. In the second experiment, to reduce complexity, the self-attention mechanism within the swin-Transformer block is replaced with an similar albeit simpler spatial mixing operation i.e. max-pooling. We observe improved performance for max-pooling in smaller dataset scales, indicating that the window-based Transformer may not be the best choice in both small and larger dataset scales. Finally, since convolution has an inherent local inductive bias of positional information, we conduct a third experiment to imbibe such a property to the Transformer by exploring two kinds of positional encodings. The results show that there are insignificant improvements after adding positional encoding, indicating the hybrid swin-Transformers deficiency in capturing positional information given our dataset at its various scales. Through this work, we hope to motivate the community to use learning curves under fair experimental settings to evaluate the efficacy of newer architectures like Transformers for their medica, Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Pattern Recognition and Bioinformatics

In radiological practice, multi-sequence MRI is routinely acquired to characterize anatomy and tissue. However, due to the heterogeneity of imaging protocols and contraindications to contrast agents, some MRI sequences, e.g. contrast-enhanced T1-weighted image (T1ce), may not be acquired. This creates difficulties for large-scale clinical studies for which heterogeneous datasets are aggregated. Modern deep learning techniques have demonstrated the capability of synthesizing missing sequences from existing sequences, through learning from an extensive multi-sequence MRI dataset. In this paper, we propose a novel MR image translation solution based on local implicit neural representations. We split the available MRI sequences into local patches and assign to each patch a local multi-layer perceptron (MLP) that represents a patch in the T1ce. The parameters of these local MLPs are generated by a hypernetwork based on image features. Experimental results and ablation studies on the BraTS challenge dataset showed that the local MLPs are critical for recovering fine image and tumor details, as they allow for local specialization that is highly important for accurate image translation. Compared to a classical pix2pix model, the proposed method demonstrated visual improvement and significantly improved quantitative scores (MSE 0.86 × 10-3 vs. 1.02 × 10-3 and SSIM 94.9 vs 94.3)., Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Pattern Recognition and Bioinformatics, ImPhys/Tao group

Purpose: Parallel RF transmission (PTx) is one of the key technologies enabling high quality imaging at ultra-high fields (≥7T). Compliance with regulatory limits on the local specific absorption rate (SAR) typically involves over-conservative safety margins to account for intersubject variability, which negatively affect the utilization of ultra-high field MR. In this work, we present a method to generate a subject-specific body model from a single T1-weighted dataset for personalized local SAR prediction in PTx neuroimaging at 7T. Methods: Multi-contrast data were acquired at 7T (N = 10) to establish ground truth segmentations in eight tissue types. A 2.5D convolutional neural network was trained using the T1-weighted data as input in a leave-one-out cross-validation study. The segmentation accuracy was evaluated through local SAR simulations in a quadrature birdcage as well as a PTx coil model. Results: The network-generated segmentations reached Dice coefficients of 86.7% ± 6.7% (mean ± SD) and showed to successfully address the severe intensity bias and contrast variations typical to 7T. Errors in peak local SAR obtained were below 3.0% in the quadrature birdcage. Results obtained in the PTx configuration indicated that a safety margin of 6.3% ensures conservative local SAR estimates in 95% of the random RF shims, compared to an average overestimation of 34% in the generic “one-size-fits-all” approach. Conclusion: A subject-specific body model can be automatically generated from a single T1-weighted dataset by means of deep learning, providing the necessary inputs for accurate and personalized local SAR predictions in PTx neuroimaging at 7T., Circuits and Systems

For image-guided small animal irradiations, the whole workflow of imaging, organ contouring, irradiation planning, and delivery is typically performed in a single session requiring continuous administration of anaesthetic agents. Automating contouring leads to a faster workflow, which limits exposure to anaesthesia and thereby, reducing its impact on experimental results and on animal wellbeing. Here, we trained the 2D and 3D U-Net architectures of no-new-Net (nnU-Net) for autocontouring of the thorax in mouse micro-CT images. We trained the models only on native CTs and evaluated their performance using an independent testing dataset (i.e., native CTs not included in the training and validation). Unlike previous studies, we also tested the model performance on an external dataset (i.e., contrast-enhanced CTs) to see how well they predict on CTs completely different from what they were trained on. We also assessed the interobserver variability using the generalized conformity index (CI gen) among three observers, providing a stronger human baseline for evaluating automated contours than previous studies. Lastly, we showed the benefit on the contouring time compared to manual contouring. The results show that 3D models of nnU-Net achieve superior segmentation accuracy and are more robust to unseen data than 2D models. For all target organs, the mean surface distance (MSD) and the Hausdorff distance (95p HD) of the best performing model for this task (nnU-Net 3d_fullres) are within 0.16 mm and 0.60 mm, respectively. These values are below the minimum required contouring accuracy of 1 mm for small animal irradiations, and improve significantly upon state-of-the-art 2D U-Net-based AIMOS method. Moreover, the conformity indices of the 3d_fullres model also compare favourably to the interobserver variability for all target organs, whereas the 2D models perform poorly in this regard. Importantly, the 3d_fullres model offers 98% reduction in contouring time., RST/Reactor Physics and Nuclear Materials

Numerical Analysis, Tera-Hertz Sensing, Pattern Recognition and Bioinformatics

Bayesian Neural Nets (BNN) are increasingly used for robust organ auto-contouring. Uncertainty heatmaps extracted from BNNs have been shown to correspond to inaccurate regions. To help speed up the mandatory quality assessment (QA) of contours in radiotherapy, these heatmaps could be used as stimuli to direct visual attention of clinicians to potential inaccuracies. In practice, this is non-trivial to achieve since many accurate regions also exhibit uncertainty. To influence the output uncertainty of a BNN, we propose a modified accuracy-versus-uncertainty (AvU) metric as an additional objective during model training that penalizes both accurate regions exhibiting uncertainty as well as inaccurate regions exhibiting certainty. For evaluation, we use an uncertainty-ROC curve that can help differentiate between Bayesian models by comparing the probability of uncertainty in inaccurate versus accurate regions. We train and evaluate a FlipOut BNN model on the MICCAI2015 Head and Neck Segmentation challenge dataset and on the DeepMind-TCIA dataset, and observed an increase in the AUC of uncertainty-ROC curves by 5.6% and 5.9%, respectively, when using the AvU objective. The AvU objective primarily reduced false positives regions (uncertain and accurate), drawing less visual attention to these regions, thereby potentially improving the speed of error detection., Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Computer Graphics and Visualisation, Pattern Recognition and Bioinformatics

Delineation of tumours and organs-at-risk permits detecting and correcting changes in the patients' anatomy throughout the treatment, making it a core step of adaptive external beam radiotherapy. Although auto-contouring technologies have sped up this process, the time needed to perform the quality assessment of the generated contours remains a bottleneck, taking clinicians between several minutes and an hour to complete. The authors of this article conducted several interviews and an observational study at two treatment centres in the Netherlands to identify challenges and opportunities for speeding up the delineation process in adaptive therapies. The study revealed three contextual variables that influence contouring performance: usable additional information, applicable domain-specific knowledge, and available editing capabilities in contouring software. In practice, clinicians leverage these variables to accelerate contouring in two ways. First, they use domain-specific knowledge and relevant clinical features such as the proximity of the organs-at-risk to the tumour to enable targeted inspection of the delineation. Second, clinicians modulate editing precision depending on the effect they anticipate the edit will have on the patient outcome. By implementing these acceleration strategies in guidelines and contouring tools, developers and workflow builders could increase contouring efficiency and consistency without affecting the patient outcome., Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Computer Graphics and Visualisation, Pattern Recognition and Bioinformatics, Human Information Communication Design

Manual or automatic delineation of the esophageal tumor in CT images is known to be very challenging. This is due to the low contrast between the tumor and adjacent tissues, the anatomical variation of the esophagus, as well as the occasional presence of foreign bodies (e.g. feeding tubes). Physicians therefore usually exploit additional knowledge such as endoscopic findings, clinical history, additional imaging modalities like PET scans. Achieving his additional information is time-consuming, while the results are error-prone and might lead to non-deterministic results. In this paper we aim to investigate if and to what extent a simplified clinical workflow based on CT alone, allows one to automatically segment the esophageal tumor with sufficient quality. For this purpose, we present a fully automatic end-to-end esophageal tumor segmentation method based on convolutional neural networks (CNNs). The proposed network, called Dilated Dense Attention Unet (DDAUnet), leverages spatial and channel attention gates in each dense block to selectively concentrate on determinant feature maps and regions. Dilated convolutional layers are used to manage GPU memory and increase the network receptive field. We collected a dataset of 792 scans from 288 distinct patients including varying anatomies with air pockets, feeding tubes and proximal tumors. Repeatability and reproducibility studies were conducted for three distinct splits of training and validation sets. The proposed network achieved a DSC value of 0.79 ± 0.20, a mean surface distance of 5.4 ± 20.2mm and 95% Hausdorff distance of 14.7 ± 25.0mm for 287 test scans, demonstrating promising results with a simplified clinical workflow based on CT alone. Our code is publicly available via https://github.com/yousefis/DenseUnet_Esophagus_Segmentation., Pattern Recognition and Bioinformatics

With the Marchenko method it is possible to retrieve Green's functions between virtual sources in the subsurface and receivers at the surface from reflection data at the surface and focusing functions. A macro model of the subsurface is needed to estimate the first arrival; the internal multiples are retrieved entirely from the reflection data. The retrieved Green's functions form the input for redatuming by multidimensional deconvolution (MDD). The redatumed reflection response is free of internal multiples related to the overburden. Alternatively, the redatumed response can be obtained by applying a second focusing function to the retrieved Green's functions. This process is called Marchenko redatuming by double focusing. It is more stable and better suited for an adaptive implementation than Marchenko redatuming by MDD, but it does not eliminate the multiples between the target and the overburden. An attractive efficient alternative is plane-wave Marchenko redatuming, which retrieves the responses to a limited number of plane-wave sources at the redatuming level. In all cases, an image of the subsurface can be obtained from the redatumed data, free of artefacts caused by internal multiples. Another class of Marchenko methods aims at eliminating the internal multiples from the reflection data, while keeping the sources and receivers at the surface. A specific characteristic of this form of multiple elimination is that it predicts and subtracts all orders of internal multiples with the correct amplitude, without needing a macro subsurface model. Like Marchenko redatuming, Marchenko multiple elimination can be implemented as an MDD process, a double dereverberation process, or an efficient plane-wave oriented process. We systematically discuss the different approaches to Marchenko redatuming, imaging and multiple elimination, using a common mathematical framework., Accepted Author Manuscript, ImPhys/Medical Imaging, Applied Geophysics and Petrophysics

In this paper we propose a supervised method to predict registration misalignment using convolutional neural networks (CNNs). This task is casted to a classification problem with multiple classes of misalignment: 'correct' 0-3 mm, 'poor' 3-6 mm and 'wrong' over 6 mm. Rather than a direct prediction, we propose a hierarchical approach, where the prediction is gradually refined from coarse to fine. Our solution is based on a convolutional Long Short-Term Memory (LSTM), using hierarchical misalignment predictions on three resolutions of the image pair, leveraging the intrinsic strengths of an LSTM for this problem. The convolutional LSTM is trained on a set of artificially generated image pairs obtained from artificial displacement vector fields (DVFs). Results on chest CT scans show that incorporating multi-resolution information, and the hierarchical use via an LSTM for this, leads to overall better F1 scores, with fewer misclassifications in a well-tuned registration setup. The final system yields an accuracy of 87.1%, and an average F1 score of 66.4% aggregated in two independent chest CT scan studies., Pattern Recognition and Bioinformatics

Suppression of surface-related and internal multiples is an outstanding challenge in seismic data processing. The former is particularly difficult in shallow water, whereas the latter is problematic for targets buried under complex, highly scattering overburdens. We have developed a two-step, amplitude- and phase-preserving, inversion-based workflow that addresses these problems. We apply robust estimation of primaries by sparse inversion (R-EPSI) to solve simultaneously for the surface-related primaries Green’s function and the source wavelet. A significant advantage of the inversion approach of the R-EPSI method is that it does not rely on an adaptive subtraction step that typically limits other demultiple methods such as surface-related multiple elimination. The resulting Green’s function is used as the input to a Marchenko equation-based approach to predict the complex interference pattern of all overburden-generated internal multiples at once. In this approach, no a priori information about the subsurface is needed. In theory, the interbed multiples can be predicted with correct amplitude and phase and, again, no adaptive filters are required. We illustrate this workflow by applying it on an Arabian Gulf field data example. It is crucial that all preprocessing steps are performed in an amplitude-preserving way to restrict any impact on the accuracy of the multiple prediction. In practice, some minor inaccuracies in the processing flow may end up as prediction errors for which corrections will be needed. Hence, we conclude that the use of conservative adaptive filters were necessary to obtain the best results after interbed multiple removal. The obtained results indicate promising suppression of surface-related and interbed multiples., Accepted Author Manuscript, Applied Geophysics and Petrophysics, ImPhys/Medical Imaging

Since the introduction of the Marchenko method in geophysics, many variants have been developed. Using a compact unified notation, we review redatuming by multidimensional deconvolution and by double focusing, virtual seismology, double dereverberation and transmission-compensated Marchenko multiple elimination, and discuss the underlying assumptions, merits and limitations of these methods., Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Applied Geophysics and Petrophysics

We have seen many developments in Marchenko equation-based methods for internal multiple attenuation in the past years. Starting from a wave-equation based method that required a smooth velocity model, there are now Marchenko equation-based methods that do not require any model information or user-input. In principle, these methods accurately predict internal multiples. Therefore, the role of the adaptive filter has changed for these methods. Rather than needing an aggressive adaptive filter to compensate for inaccurate internal multiple predictions, only a conservative adaptive filter is needed to compensate for minor amplitude and/or phase errors in the internal multiple predictions caused by imperfect acquisition and preprocessing of the input data. We demonstate that a conservative adaptive filter can be used to improve the attenuation of internal multiples when applying a Marchenko multiple elimination (MME) method to a 2D line of streamer data. In addition, we suggest that an adaptive filter can be used as a feedback mechanism to improve the preprocessing of the input data., Accepted Author Manuscript, Applied Geophysics and Petrophysics

Curiosity regarding what we cannot see has always driven research. Science has helped us to uncover many of those hidden secrets. In particular, geophysics has helped us to image the inside of the Earth. By sending a seismic signal into the Earth and recording the signal that comes back, geophysicists can characterize the layers of the subsurface. Nowadays, geophysics is used for many purposes, for example, the localization of fossil fuels, the characterization of the subsurface for the construction of wind farms and the evaluation of reservoirs for geothermal energy. In order to decrease the risk and cost involved in these activities, we need images of the subsurface that are as accurate as possible. These images can only be obtained if we fully understand the propagation of the seismic signal in the subsurface. A long-standing problem in geophysical imaging is the presence of internal multiple reflections. When imaging the subsurface, we assume that the signal only reflects once when there is a contrast in velocity and/or density (for example, when changing from sand to rock). However, in reality, the signal can reflect many times inside the subsurface before being recorded at the surface. When treating the arrivals that have reflected many times as arrivals that have only reflected once, we incorrectly image the subsurface and create ghost reflectors that do not exist. This problem is particularly strong in geological settings that have a complex structure with many strong velocity and/or density contrasts above an area of interest. This may happen, for example, when there is a reservoir of oil below a thick stratified salt layer. In such cases, the image of the area of interest is unreliable due to the presence of many ghost reflectors. Therefore, we have to use knowledge of wave propagation to predict and attenuate the internal multiples in the data prior to imaging. In this thesis, I further develop the data-driven and wave-equation-based Ma, Applied Geophysics and Petrophysics

In recent years, a variety of Marchenko methods for the attenuation of internal multiples has been developed. These methods have been extensively tested on two-dimensional synthetic data and applied to two-dimensional field data, but only little is known about their behaviour on three-dimensional synthetic data and three-dimensional field data. Particularly, it is not known whether Marchenko methods are sufficiently robust for sparse acquisition geometries that are found in practice. Therefore, we start by performing a series of synthetic tests to identify the key acquisition parameters and limitations that affect the result of three-dimensional Marchenko internal multiple prediction and subtraction using an adaptive double-focusing method. Based on these tests, we define an interpolation strategy and use it for the field data application. Starting from a wide azimuth dense grid of sources and receivers, a series of decimation tests are performed until a narrow azimuth streamer geometry remains. We evaluate the effect of the removal of sail lines, near offsets, far offsets and outer cables on the result of the adaptive double-focusing method. These tests show that our method is most sensitive to the limited aperture in the crossline direction and the sail line spacing when applying it to synthetic narrow azimuth streamer data. The sail line spacing can be interpolated, but the aperture in the crossline direction is a limitation of the acquisition. Next, we apply the adaptive Marchenko double-focusing method to the narrow azimuth streamer field data from the Santos Basin, Brazil. Internal multiples are predicted and adaptively subtracted, thereby improving the geological interpretation of the target area. These results imply that our adaptive double-focusing method is sufficiently robust for the application to three-dimensional field data, although the key acquisition parameters and limitations will naturally differ in other geological settings and for other types, Applied Geophysics and Petrophysics, ImPhys/Medical Imaging

Adaptive intelligence aims at empowering machine learning techniques with the additional use of domain knowledge. In this work, we present the application of adaptive intelligence to accelerate MR acquisition. Starting from undersampled k-space data, an iterative learning-based reconstruction scheme inspired by compressed sensing theory is used to reconstruct the images. We developed a novel deep neural network to refine and correct prior reconstruction assumptions given the training data. The network was trained and tested on a knee MRI dataset from the 2019 fastMRI challenge organized by Facebook AI Research and NYU Langone Health. All submissions to the challenge were initially ranked based on similarity with a known groundtruth, after which the top 4 submissions were evaluated radiologically. Our method was evaluated by the fastMRI organizers on an independent challenge dataset. It ranked #1, shared #1, and #3 on respectively the 8× accelerated multi-coil, the 4× multi-coil, and the 4× single-coil tracks. This demonstrates the superior performance and wide applicability of the method., Pattern Recognition and Bioinformatics

Purpose To develop and validate a robust and accurate registration pipeline for automatic contour propagation for online adaptive Intensity‐Modulated Proton Therapy (IMPT) of prostate cancer using elastix software and deep learning. Methods A three‐dimensional (3D) Convolutional Neural Network was trained for automatic bladder segmentation of the computed tomography (CT) scans. The automatic bladder segmentation alongside the computed tomography (CT) scan is jointly optimized to add explicit knowledge about the underlying anatomy to the registration algorithm. We included three datasets from different institutes and CT manufacturers. The first was used for training and testing the ConvNet, where the second and the third were used for evaluation of the proposed pipeline. The system performance was quantified geometrically using the dice similarity coefficient (DSC), the mean surface distance (MSD), and the 95% Hausdorff distance (HD). The propagated contours were validated clinically through generating the associated IMPT plans and compare it with the IMPT plans based on the manual delineations. Propagated contours were considered clinically acceptable if their treatment plans met the dosimetric coverage constraints on the manual contours. Results The bladder segmentation network achieved a DSC of 88% and 82% on the test datasets. The proposed registration pipeline achieved a MSD of 1.29 ± 0.39, 1.48 ± 1.16, and 1.49 ± 0.44 mm for the prostate, seminal vesicles, and lymph nodes, respectively, on the second dataset and a MSD of 2.31 ± 1.92 and 1.76 ± 1.39 mm for the prostate and seminal vesicles on the third dataset. The automatically propagated contours met the dose coverage constraints in 86%, 91%, and 99% of the cases for the prostate, seminal vesicles, and lymph nodes, respectively. A Conservative Success Rate (CSR) of 80% was obtained, compared to 65% when only using intensity‐based registration. Conclusion The proposed registration pipeline obtained highly promi

We apply Marchenko redatuming using an adaptive double-focusing method to 3D field data of the Santos basin, Brazil. This method was already successfully applied to 2D field data and we now study the acquisition geometry and preprocessing requirements in 3D. We start from 3D synthetic data modeled on a dense grid of colocated sources and receivers and decimate down to a realistic NAZ streamer acquisition. The synthetic tests show that the sail line spacing and the missing outer cables are the acquisition parameters with the strongest effect on Marchenko redatuming. We can interpolate for the sail line spacing and the near offsets, but the missing outer cables are unfortunately a limitation of the acquisition. After applying the proposed interpolation to 3D field data, interbed multiples are successfully predicted and subtracted from the target area, resulting in a significant improvement in the geological interpretation. Naturally, the pre-processing requirements and challenges strongly depend on the acquisition geometry and the geology of the area under investigation (e.g. water depth, shape of the overburden, maximum dip). Hence, these tests only give a general idea about the limitations of 3D Marchenko redatuming, Accepted author manuscript, Applied Geophysics and Petrophysics, ImPhys/Acoustical Wavefield Imaging

Objective: Our goal was to investigate the performance of an open source deformable image registration package, elastix, for fast and robust contour propagation in the context of online-adaptive intensity-modulated proton therapy (IMPT) for prostate cancer. Methods: A planning and 7–10 repeat CT scans were available of 18 prostate cancer patients. Automatic contour propagation of repeat CT scans was performed using elastix and compared with manual delineations in terms of geometric accuracy and runtime. Dosimetric accuracy was quantified by generating IMPT plans using the propagated contours expanded with a 2 mm (prostate) and 3.5 mm margin (seminal vesicles and lymph nodes) and calculating dosimetric coverage based on the manual delineation. A coverage of V95% ≥ 98% (at least 98% of the target volumes receive at least 95% of the prescribed dose) was considered clinically acceptable. Results: Contour propagation runtime varied between 3 and 30 s for different registration settings. For the fastest setting, 83 in 93 (89.2%), 73 in 93 (78.5%), and 91 in 93 (97.9%) registrations yielded clinically acceptable dosimetric coverage of the prostate, seminal vesicles, and lymph nodes, respectively. For the prostate, seminal vesicles, and lymph nodes the Dice Similarity Coefficient (DSC) was 0.87 ± 0.05, 0.63 ± 0.18, and 0.89 ± 0.03 and the mean surface distance (MSD) was 1.4 ± 0.5 mm, 2.0 ± 1.2 mm, and 1.5 ± 0.4 mm, respectively. Conclusion: With a dosimetric success rate of 78.5–97.9%, this software may facilitate online adaptive IMPT of prostate cancer using a fast, free and open implementation., Pattern Recognition and Bioinformatics

The problem of motion detection has received considerable attention due to the explosive growth of its applications in video analysis and surveillance systems. While the previous approaches can produce good results, the accurate detection of motion remains a challenging task due to the difficulties raised by illumination variations, occlusion, camouflage, sudden motions appearing in burst, dynamic texture, and environmental changes such as those on weather conditions, sunlight changes during a day, etc. In this study, a novel per-pixel motion descriptor is proposed for motion detection in video sequences which outperforms the current methods in the literature particularly in severe scenarios. The proposed descriptor is based on two complementary three-dimensional discrete wavelet transforms (3D-DWT) and a three-dimensional wavelet leader. In this approach, a feature vector is extracted for each pixel by applying a novel three-dimensional wavelet-based motion descriptor. Then, the extracted features are clustered by the well-known K-means algorithm. The experimental results demonstrate the effectiveness of the proposed method compared to state-of-the-art approaches in several public benchmark datasets. The application of the proposed method and additional experimental results for several challenging datasets are available online., This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication., Pattern Recognition and Bioinformatics

The Santos basin in Brazil suffers from strong internal multiples that overlap with primaries from the pre-salt reservoirs. We propose an adaptive double-focusing method for the removal of these multiples to obtain a correct image of the target area. The proposed method applies a form of source-receiver Marchenko redatuming to the reflection response. The Marchenko method is used to achieve single-focusing, after which we convolve the retrieved downgoing focusing function and the upgoing Green’s function to create double-focusing. This results in a base image that contains both primaries and internal multiples, and two models that predict the strongest internal multiples. Next, adaptive subtraction in the curvelet domain is used to remove these multiples from the base image. Some multiple interactions between the target area and the overburden remain, but we gain a robust method that is capable of dealing with a sparse acquisition geometry and imperfections in the (pre-processed) data. Also, this method is straightforward to implement and can be parallelized over pairs of focal points. These properties make adaptive double-focusing particularly suitable for the application to large volumes of field data. Tests on 2D field data and 3D field data show that the proposed method correctly predicts and removes the strongest internal multiples from the overburden, resulting in a clear improvement of the geological interpretability in the target area., Accepted author manuscript, Applied Geophysics and Petrophysics, ImPhys/Acoustical Wavefield Imaging

In seismic monitoring, one is usually interested in the response of a changing target zone, embedded in a static inhomogeneous medium. We introduce an efficient method that predicts reflection responses at the Earth's surface for different target-zone scenarios, from a single reflection response at the surface and a model of the changing target zone. The proposed process consists of two main steps. In the first step, the response of the original target zone is removed from the reflection response, using the Marchenko method. In the second step, the modelled response of a new target zone is inserted between the overburden and underburden responses. The method fully accounts for all orders of multiple scattering and, in the elastodynamic case, for wave conversion. For monitoring purposes, only the second step needs to be repeated for each target-zone model. Since the target zone covers only a small part of the entire medium, the proposed method is much more efficient than repeated modelling of the entire reflection response., ImPhys/Acoustical Wavefield Imaging, Applied Geophysics and Petrophysics

Internal multiples can create severe artefacts in seismic imaging, especially when our zone of interest is overlain by a complex overburden. These artefacts can mask structures, which has a strong effect on the interpretation of the image. Therefore, multiple prediction and removal is of significant importance for correct imaging and interpretation in settings with a complex overburden. We propose an adaptive double-focusing method to predict and subtract the internal multiples that were generated in the overburden. This method is a form of the Marchenko method, that can retrieve the directionally-decomposed Green's functions between virtual sources and virtual receivers anywhere inside the subsurface. The retrieved Green's functions contain all orders of multiple scattering. The method only requires the single-sided reflection response and a smooth velocity model as input. Instead of conventional imaging methods, that assume that the wavefield only consists of single-scattered waves (and thus create imaging artefacts when multiple scattering is present), we now use the multiple-scattered Marchenko wavefields for correct redatuming and imaging. We apply our method to 2D and 3D field data that were recorded in settings where imaging and interpretation is hindered by a complex overburden. First, we create virtual sources and virtual receivers directly above our zone of interest. Next, we use the retrieved Marchenko wavefields to predict and subtract the internal multiples that were generated in the overburden. Masked structures become visible after multiple removal, which significantly improves the geological interpretability. Therefore, we conclude that the adaptive double-focusing method (Marchenko redatuming) is capable of correctly predicting and removing internal multiples generated in the overburden., S24A-03 presented at 2018 Fall Meeting, AGU, Washington, D.C., 10-14 Dec. Session: [S24A] Frontiers in Theoretical and Computational Seismology I, ImPhys/Acoustical Wavefield Imaging, Applied Geophysics and Petrophysics

Recent developments in exploration seismology have enabled the creation of virtual sources and/or virtual receivers in the subsurface from reflection measurements at the earth's surface. Unlike in seismic interferometry, no physical instrument (receiver or source) is needed at the position of the virtual source or receiver. Moreover, no detailed knowledge of the subsurface parameters and structures is required: a smooth velocity model suffices. Yet, the responses to the virtual sources, observed by the virtual receivers, fully account for multiple scattering. This new methodology, which we call virtual seismology, has led to a breakthrough in hydrocarbon reservoir imaging, as is demonstrated in a companion paper (Staring et al., Marchenko redatuming for multiple prediction and removal in situations with a complex overburden). The aim of the present paper is to discuss applications of virtual seismology beyond exploration seismology, in particular induced earthquake monitoring, and to highlight the connections between these applications. The ability to retrieve the entire wave field between (virtual or real) sources and receivers anywhere in the subsurface, without needing a detailed subsurface model, has large potential for monitoring induced seismicity, characterizing the source properties (such as the moment tensor of extended sources along a fault plane), and forecasting the response to potential future induced earthquakes. This will be demonstrated with numerical models and preliminary real-data results., S53A-03 presented at 2018 Fall Meeting, AGU, Washington, D.C., 10-14 Dec. Session: S53A On the Symbiosis Between Fundamental and Exploration Geophysics I, ImPhys/Acoustical Wavefield Imaging, Applied Geophysics and Petrophysics

In recent years, machine learning approaches have been successfully applied to the field of neuroimaging for classification and regression tasks. However, many approaches do not give an intuitive relation between the raw features and the diagnosis. Therefore, they are difficult for clinicians to interpret. Moreover, most approaches treat the features extracted from the brain (for example, voxelwise gray matter concentration maps from brain MRI) as independent variables and ignore their spatial and anatomical relations. In this paper, we present a new Support Vector Machine (SVM)-based learning method for the classification of Alzheimer's disease (AD), which integrates spatial-anatomical information. In this way, spatial-neighbor features in the same anatomical region are encouraged to have similar weights in the SVM model. Secondly, we introduce a group lasso penalty to induce structure sparsity, which may help clinicians to assess the key regions involved in the disease. For solving this learning problem, we use an accelerated proximal gradient descent approach. We tested our method on the subset of ADNI data selected by Cuingnet et al. (2011) for Alzheimer's disease classification, as well as on an independent larger dataset from ADNI. Good classification performance is obtained for distinguishing cognitive normals (CN) vs. AD, as well as on distinguishing between various sub-types (e.g. CN vs. Mild Cognitive Impairment). The model trained on Cuignet's dataset for AD vs. CN classification was directly used without re-training to the independent larger dataset. Good performance was achieved, demonstrating the generalizability of the proposed methods. For all experiments, the classification results are comparable or better than the state-of-the-art, while the weight map more clearly indicates the key regions related to Alzheimer's disease., Pattern Recognition and Bioinformatics

We have developed an adaptive double-focusing method that is specifically designed for the field-data application of source-receiver Marchenko redatuming. Typically, the single-focusing Marchenko method is combined with a multidimensional deconvolution (MDD) to achieve redatuming. Our method replaces the MDD step by a second focusing step that naturally complements the single-focusing Marchenko method. Instead of performing the MDD method with the directionally decomposed Green's functions that result from single-focusing, we now use the retrieved upgoing Green's function and the retrieved downgoing focusing function to obtain a redatumed reflection response in the physical medium. Consequently, we only remove the strongest overburden effects instead of removing all of the overburden effects. However, the gain is a robust method that is less sensitive to imperfections in the data and a sparse acquisition geometry than the MDD method. In addition, it is computationally much cheaper, more straightforward to implement, and it can be parallelized over pairs of focal points, which makes it suitable for application to large data volumes. We evaluate the successful application of our method to 2D field data of the Santos Basin., Applied Geophysics and Petrophysics, ImPhys/Acoustical Wavefield Imaging