Search

In recent years, the increasing size of deep learning models and their growing demand for computational resources have drawn significant attention to the practice of pruning neural networks, while aiming to preserve their accuracy. In unstructured gradual pruning, which sparsifies a network by gradually removing individual network parameters until a targeted network sparsity is reached, recent works show that both gradient and weight magnitudes should be considered. In this work, we show that such mechanism, e.g., the order of prioritization and selection criteria, is essential. We introduce a gradient-first magnitude-next strategy for choosing the parameters to prune, and show that a fixed-rate subselection criterion between these steps works better, in contrast to the annealing approach in the literature. We validate this on CIFAR-10 dataset, with multiple randomized initializations on both VGG-19 and ResNet-50 network backbones, for pruning targets of 90, 95, and 98% sparsity and for both initially dense and 50% sparse networks. Our proposed fixed-rate gradient-first gradual pruning (FGGP) approach outperforms its state-of-the-art alternatives in most of the above experimental settings, even occasionally surpassing the upperbound of corresponding dense network results, and having the highest ranking across the considered experimental settings.

Despite their effectiveness, current deep learning models face challenges with images coming from different domains with varying appearance and content. We introduce SegCLR, a versatile framework designed to segment volumetric images across different domains, employing supervised and contrastive learning simultaneously to effectively learn from both labeled and unlabeled data. We demonstrate the superior performance of SegCLR through a comprehensive evaluation involving three diverse clinical datasets of retinal fluid segmentation in 3D Optical Coherence Tomography (OCT), various network configurations, and verification across 10 different network initializations. In an unsupervised domain adaptation context, SegCLR achieves results on par with a supervised upper-bound model trained on the intended target domain. Notably, we discover that the segmentation performance of SegCLR framework is marginally impacted by the abundance of unlabeled data from the target domain, thereby we also propose an effective zero-shot domain adaptation extension of SegCLR, eliminating the need for any target domain information. This shows that our proposed addition of contrastive loss in standard supervised training for segmentation leads to superior models, inherently more generalizable to both in- and out-of-domain test data. We additionally propose a pragmatic solution for SegCLR deployment in realistic scenarios with multiple domains containing labeled data. Accordingly, our framework pushes the boundaries of deep-learning based segmentation in multi-domain applications, regardless of data availability - labeled, unlabeled, or nonexistent.

Speed-of-sound (SoS) is an emerging ultrasound contrast modality, where pulse-echo techniques using conventional transducers offer multiple benefits. For estimating tissue SoS distributions, spatial domain reconstruction from relative speckle shifts between different beamforming sequences is a promising approach. This operates based on a forward model that relates the sought local values of SoS to observed speckle shifts, for which the associated image reconstruction inverse problem is solved. The reconstruction accuracy thus highly depends on the hand-crafted forward imaging model. In this work, we propose to learn the SoS imaging model based on data. We introduce a convolutional formulation of the pulse-echo SoS imaging problem such that the entire field-of-view requires a single unified kernel, the learning of which is then tractable and robust. We present least-squares estimation of such convolutional kernel, which can further be constrained and regularized for numerical stability. In experiments, we show that a forward model learned from k-Wave simulations improves the median contrast of SoS reconstructions by 63%, compared to a conventional hand-crafted line-based wave-path model. This simulation-learned model generalizes successfully to acquired phantom data, nearly doubling the SoS contrast compared to the conventional hand-crafted alternative. We demonstrate equipment-specific and small-data regime feasibility by learning a forward model from a single phantom image, where our learned model quadruples the SoS contrast compared to the conventional hand-crafted model. On in-vivo data, the simulation- and phantom-learned models respectively exhibit impressive 7 and 10 folds contrast improvements over the conventional model.

Speed-of-sound (SoS) is a novel imaging biomarker for assessing biomechanical characteristics of soft tissues. SoS imaging in pulse-echo mode using conventional ultrasound systems with hand-held transducers has the potential to enable new clinical uses. Recent work demonstrated diverging waves from single-element (SE) transmits to outperform plane-wave sequences. However, single-element transmits have severely limited power and hence produce low signal-to-noise ratio (SNR) in echo data. We herein propose Walsh-Hadamard (WH) coded and virtual-source (VS) transmit sequences for improved SNR in SoS imaging. We additionally present an iterative method of estimating beamforming SoS in the medium, which otherwise confound SoS reconstructions due to beamforming inaccuracies in the images used for reconstruction. Through numerical simulations, phantom experiments, and in-vivo imaging data, we show that WH is not robust against motion, which is often unavoidable in clinical imaging scenarios. Our proposed virtual-source sequence is shown to provide the highest SoS reconstruction performance, especially robust to motion artifacts. In phantom experiments, despite having a comparable SoS root-mean-square-error (RMSE) of 17.5 to 18.0 m/s at rest, with a minor axial probe motion of ~0.67 mm/s the RMSE for SE, WH, and VS already deteriorate to 20.2, 105.4, 19.0 m/s, respectively; showing that WH produces unacceptable results, not robust to motion. In the clinical data, the high SNR and motion-resilience of VS sequence is seen to yield superior contrast compared to SE and WH sequences.

Photorealistic image generation from simulated label maps are necessitated in several contexts, such as for medical training in virtual reality. With conventional deep learning methods, this task requires images that are paired with semantic annotations, which typically are unavailable. We introduce a contrastive learning framework for generating photorealistic images from simulated label maps, by learning from unpaired sets of both. Due to potentially large scene differences between real images and label maps, existing unpaired image translation methods lead to artifacts of scene modification in synthesized images. We utilize simulated images as surrogate targets for a contrastive loss, while ensuring consistency by utilizing features from a reverse translation network. Our method enables bidirectional label-image translations, which is demonstrated in a variety of scenarios and datasets, including laparoscopy, ultrasound, and driving scenes. By comparing with state-of-the-art unpaired translation methods, our proposed method is shown to generate realistic and scene-accurate translations.

Methods for unsupervised domain adaptation (UDA) help to improve the performance of deep neural networks on unseen domains without any labeled data. Especially in medical disciplines such as histopathology, this is crucial since large datasets with detailed annotations are scarce. While the majority of existing UDA methods focus on the adaptation from a labeled source to a single unlabeled target domain, many real-world applications with a long life cycle involve more than one target domain. Thus, the ability to sequentially adapt to multiple target domains becomes essential. In settings where the data from previously seen domains cannot be stored, e.g., due to data protection regulations, the above becomes a challenging continual learning problem. To this end, we propose to use generative feature-driven image replay in conjunction with a dual-purpose discriminator that not only enables the generation of images with realistic features for replay, but also promotes feature alignment during domain adaptation. We evaluate our approach extensively on a sequence of three histopathological datasets for tissue-type classification, achieving state-of-the-art results. We present detailed ablation experiments studying our proposed method components and demonstrate a possible use-case of our continual UDA method for an unsupervised patch-based segmentation task given high-resolution tissue images.

The segmentation and automatic identification of histological regions of diagnostic interest offer a valuable aid to pathologists. However, segmentation methods are hampered by the difficulty of obtaining pixel-level annotations, which are tedious and expensive to obtain for Whole-Slide images (WSI). To remedy this, weakly supervised methods have been developed to exploit the annotations directly available at the image level. However, to our knowledge, none of these techniques is adapted to deal with WSIs. In this paper, we propose WholeSIGHT, a weakly-supervised method, to simultaneously segment and classify WSIs of arbitrary shapes and sizes. Formally, WholeSIGHT first constructs a tissue-graph representation of the WSI, where the nodes and edges depict tissue regions and their interactions, respectively. During training, a graph classification head classifies the WSI and produces node-level pseudo labels via post-hoc feature attribution. These pseudo labels are then used to train a node classification head for WSI segmentation. During testing, both heads simultaneously render class prediction and segmentation for an input WSI. We evaluated WholeSIGHT on three public prostate cancer WSI datasets. Our method achieved state-of-the-art weakly-supervised segmentation performance on all datasets while resulting in better or comparable classification with respect to state-of-the-art weakly-supervised WSI classification methods. Additionally, we quantify the generalization capability of our method in terms of segmentation and classification performance, uncertainty estimation, and model calibration.

Deep learning models can achieve high accuracy when trained on large amounts of labeled data. However, real-world scenarios often involve several challenges: Training data may become available in installments, may originate from multiple different domains, and may not contain labels for training. Certain settings, for instance medical applications, often involve further restrictions that prohibit retention of previously seen data due to privacy regulations. In this work, to address such challenges, we study unsupervised segmentation in continual learning scenarios that involve domain shift. To that end, we introduce GarDA (Generative Appearance Replay for continual Domain Adaptation), a generative-replay based approach that can adapt a segmentation model sequentially to new domains with unlabeled data. In contrast to single-step unsupervised domain adaptation (UDA), continual adaptation to a sequence of domains enables leveraging and consolidation of information from multiple domains. Unlike previous approaches in incremental UDA, our method does not require access to previously seen data, making it applicable in many practical scenarios. We evaluate GarDA on two datasets with different organs and modalities, where it substantially outperforms existing techniques., Comment: Fixed typos

Most ultrasound imaging techniques necessitate the fundamental step of converting temporal signals received from transducer elements into a spatial echogenecity map. This beamforming (BF) step requires the knowledge of speed-of-sound (SoS) value in the imaged medium. An incorrect assumption of BF SoS leads to aberration artifacts, not only deteriorating the quality and resolution of conventional brightness mode (B-mode) images, hence limiting their clinical usability, but also impairing other ultrasound modalities such as elastography and spatial SoS reconstructions, which rely on faithfully beamformed images as their input. In this work, we propose an analytical method for estimating BF SoS. We show that pixel-wise relative shifts between frames beamformed with an assumed SoS is a function of geometric disparities of the transmission paths and the error in such SoS assumption. Using this relation, we devise an analytical model, the closed form solution of which yields the difference between the assumed and the true SoS in the medium. Based on this, we correct the BF SoS, which can also be applied iteratively. Both in simulations and experiments, lateral B-mode resolution is shown to be improved by $\approx$25% compared to that with an initial SoS assumption error of 3.3% (50 m/s), while localization artifacts from beamforming are also corrected. After 5 iterations, our method achieves BF SoS errors of under 0.6 m/s in simulations and under 1 m/s in experimental phantom data. Residual time-delay errors in beamforming 32 numerical phantoms are shown to reduce down to 0.07 $\mu$s, with average improvements of up to 21 folds compared to initial inaccurate assumptions. We additionally show the utility of the proposed method in imaging local SoS maps, where using our correction method reduces reconstruction root-mean-square errors substantially, down to their lower-bound with actual BF SoS.

Most ultrasound (US) imaging techniques use spatially-constant speed-of-sound (SoS) values for beamforming. Having a discrepancy between the actual and used SoS value leads to aberration artifacts, e.g., reducing the image resolution, which may affect diagnostic usability. Accuracy and quality of different US imaging modalities, such as tomographic reconstruction of local SoS maps, also depend on a good initial beamforming SoS. In this work, we develop an analytical method for estimating mean SoS in an imaged medium. We show that the relative shifts between beamformed frames depend on the SoS offset and the geometric disparities in transmission paths. Using this relation, we estimate a correction factor and hence a corrected mean SoS in the medium. We evaluated our proposed method on a set of numerical simulations, demonstrating its utility both for global SoS prediction and for local SoS tomographic reconstruction. For our evaluation dataset, for an initial SoS under- and over-assumption of 5% the medium SoS, our method is able to predict the actual mean SoS within 0.3% accuracy. For the tomographic reconstruction of local SoS maps, the reconstruction accuracy is improved on average by 78.5% and 87%, respectively, compared to an initial SoS under- and over-assumption of 5%.

Multiple Instance Learning (MIL) methods have become increasingly popular for classifying giga-pixel sized Whole-Slide Images (WSIs) in digital pathology. Most MIL methods operate at a single WSI magnification, by processing all the tissue patches. Such a formulation induces high computational requirements, and constrains the contextualization of the WSI-level representation to a single scale. A few MIL methods extend to multiple scales, but are computationally more demanding. In this paper, inspired by the pathological diagnostic process, we propose ZoomMIL, a method that learns to perform multi-level zooming in an end-to-end manner. ZoomMIL builds WSI representations by aggregating tissue-context information from multiple magnifications. The proposed method outperforms the state-of-the-art MIL methods in WSI classification on two large datasets, while significantly reducing the computational demands with regard to Floating-Point Operations (FLOPs) and processing time by up to 40x., Comment: Typos corrected; Changed dataset name from INSEC to CRC upon dataset creators' request; Update affiliation and fix typos

Accurate segmentation of retinal fluids in 3D Optical Coherence Tomography images is key for diagnosis and personalized treatment of eye diseases. While deep learning has been successful at this task, trained supervised models often fail for images that do not resemble labeled examples, e.g. for images acquired using different devices. We hereby propose a novel semi-supervised learning framework for segmentation of volumetric images from new unlabeled domains. We jointly use supervised and contrastive learning, also introducing a contrastive pairing scheme that leverages similarity between nearby slices in 3D. In addition, we propose channel-wise aggregation as an alternative to conventional spatial-pooling aggregation for contrastive feature map projection. We evaluate our methods for domain adaptation from a (labeled) source domain to an (unlabeled) target domain, each containing images acquired with different acquisition devices. In the target domain, our method achieves a Dice coefficient 13.8% higher than SimCLR (a state-of-the-art contrastive framework), and leads to results comparable to an upper bound with supervised training in that domain. In the source domain, our model also improves the results by 5.4% Dice, by successfully leveraging information from many unlabeled images., Comment: Accepted for publication at MICCAI 2022

Speed-of-sound and attenuation of ultrasound waves vary in the tissues. There exist methods in the literature that allow for spatially reconstructing the distribution of group speed-of-sound (SoS) and frequency-dependent ultrasound attenuation (UA) using reflections from an acoustic mirror positioned at a known distance from the transducer. These methods utilize a conventional ultrasound transducer operating in pulse-echo mode and a calibration protocol with measurements in water. In this study, we introduce a novel method for reconstructing local SoS and UA maps as a function of acoustic frequency through Fourier-domain analysis and by fitting linear and power-law dependency models in closed form. Frequency-dependent SoS and UA together characterize the tissue comprehensively in spectral domain within the utilized transducer bandwidth. In simulations, our proposed methods are shown to yield low reconstruction error: 0.01 dB/cm.MHz^y for attenuation coefficient and 0.05 for the frequency exponent. For tissue-mimicking phantoms and ex-vivo bovine muscle samples, a high reconstruction contrast was achieved. Attenuation exponents in a gelatin-cellulose mixture and an ex-vivo bovine muscle sample were found to be, respectively, 1.3 and 0.6 on average. Linear dispersion of SoS in a gelatin-cellulose mixture and an ex-vivo bovine muscle sample were found to be, respectively, 1.3 and 4.0 m/s.MHz on average. These findings were reproducible when the inclusion and substrate materials were exchanged. Bulk loss modulus in the bovine muscle sample was computed to be approximately 4 times the bulk loss modulus in the gelatin-cellulose mixture. Such frequency-dependent characteristics of SoS and UA, and bulk loss modulus may therefore differentiate tissues as potential diagnostic biomarkers.

In ultrasound beamforming, focusing time delays are typically computed with a spatially constant speed-of-sound (SoS) assumption. A mismatch between beamforming and true medium SoS then leads to aberration artifacts. Other imaging techniques such as spatially-resolved SoS reconstruction using tomographic techniques also rely on a good SoS estimate for initial beamforming. In this work, we exploit spatially-varying geometric disparities in the transmit and receive paths of multiple sequences for estimating a mean medium SoS. We use images from diverging waves beamformed with an assumed SoS, and propose a model fitting method for estimating the SoS offset. We demonstrate the effectiveness of our proposed method for tomographic SoS reconstruction. With corrected beamforming SoS, the reconstruction accuracy on simulated data was improved by 63% and 29%, respectively, for an initial SoS over- and under-estimation of 1.5%. We further demonstrate our proposed method on a breast phantom, indicating substantial improvement in contrast-to-noise ratio for local SoS mapping.

Shear-wave Elastography Imaging (SWEI) is a noninvasive imaging modality that provides tissue elasticity information by measuring the travelling speed of an induced shear-wave. It is commercially available on clinical ultrasound scanners and popularly used in the diagnosis and staging of liver disease and breast cancer. In conventional SWEI methods, a sequence of acoustic radiation force (ARF) pushes are used for inducing a shear-wave, which is tracked using high frame-rate multi-angle plane wave imaging (MA-PWI) to estimate the shear-wave speed (SWS). Conventionally, these plane waves are beamformed using a constant speed-of-sound (SoS), assuming an a-priori known and homogeneous tissue medium. However, soft tissues are inhomogeneous, with intrinsic SoS variations. In this work, study the SoS effects and inhomogeneities on SWS estimation, using simulation and phantoms experiments with porcine muscle as an abbarator, and show how these aberrations can be corrected using local speed-of-sound adaptive beamforming. For shear-wave tracking, we compare standard beamform with spatially constant SoS values to software beamforming with locally varying SoS maps. We show that, given SoS aberrations, traditional beamforming using a constant SoS, regardless of the utilized SoS value, introduces a substantial bias in resulting SWS estimations. Average SWS estimation disparity for the same material was observed over 4.3 times worse when a constant SoS value is used compared to that when a known SoS map is used for beamforming. Such biases are shown to be corrected by using a local SoS map in beamforming, indicating the importance of and a need for local SoS reconstruction techniques.

Neural networks are prone to catastrophic forgetting when trained incrementally on different tasks. Popular incremental learning methods mitigate such forgetting by retaining a subset of previously seen samples and replaying them during the training on subsequent tasks. However, this is not always possible, e.g., due to data protection regulations. In such restricted scenarios, one can employ generative models to replay either artificial images or hidden features to a classifier. In this work, we propose Genifer (GENeratIve FEature-driven image Replay), where a generative model is trained to replay images that must induce the same hidden features as real samples when they are passed through the classifier. Our technique therefore incorporates the benefits of both image and feature replay, i.e.: (1) unlike conventional image replay, our generative model explicitly learns the distribution of features that are relevant for classification; (2) in contrast to feature replay, our entire classifier remains trainable; and (3) we can leverage image-space augmentations, which increase distillation performance while also mitigating overfitting during the training of the generative model. We show that Genifer substantially outperforms the previous state of the art for various settings on the CIFAR-100 and CUB-200 datasets., Comment: Revised method, evaluation, and discussion

Displacement estimation in optical coherence tomography (OCT) imaging is relevant for several potential applications, e.g. for optical coherence elastography (OCE) for corneal biomechanical characterization. Larger displacements may be resolved using correlation-based block matching techniques, which however are prone to signal de-correlation and imprecise at commonly desired sub-pixel resolutions. Phase-based tracking methods can estimate tiny sub-wavelength motion, but are not suitable for motion magnitudes larger than half the wavelength due to phase wrapping and the difficulty of any unwrapping due to noise. In this paper a robust OCT displacement estimation method is introduced by formulating tracking as an optimization problem that jointly penalizes intensity disparity, phase difference, and motion discontinuity. This is then solved using yynamic programming, utilizing both sub-wavelength-scale phase and pixel-scale intensity information from OCT imaging, while inherently seeking for the number of phase wraps. This allows for effectively tracking axial and lateral displacements, respectively, with sub-wavelength and pixel scale resolution. Results with tissue mimicking phantoms show that our proposed approach substantially outperforms conventional methods in terms of axial tracking precision, in particular for displacements exceeding half the imaging wavelength., Comment: 10 pages, 8 figures

Interactive simulation of ultrasound imaging greatly facilitates sonography training. Although ray-tracing based methods have shown promising results, obtaining realistic images requires substantial modeling effort and manual parameter tuning. In addition, current techniques still result in a significant appearance gap between simulated images and real clinical scans. Herein we introduce a novel content-preserving image translation framework (ConPres) to bridge this appearance gap, while maintaining the simulated anatomical layout. We achieve this goal by leveraging both simulated images with semantic segmentations and unpaired in-vivo ultrasound scans. Our framework is based on recent contrastive unpaired translation techniques and we propose a regularization approach by learning an auxiliary segmentation-to-real image translation task, which encourages the disentanglement of content and style. In addition, we extend the generator to be class-conditional, which enables the incorporation of additional losses, in particular a cyclic consistency loss, to further improve the translation quality. Qualitative and quantitative comparisons against state-of-the-art unpaired translation methods demonstrate the superiority of our proposed framework.

Segmenting histology images into diagnostically relevant regions is imperative to support timely and reliable decisions by pathologists. To this end, computer-aided techniques have been proposed to delineate relevant regions in scanned histology slides. However, the techniques necessitate task-specific large datasets of annotated pixels, which is tedious, time-consuming, expensive, and infeasible to acquire for many histology tasks. Thus, weakly-supervised semantic segmentation techniques are proposed to utilize weak supervision that is cheaper and quicker to acquire. In this paper, we propose SegGini, a weakly supervised segmentation method using graphs, that can utilize weak multiplex annotations, i.e. inexact and incomplete annotations, to segment arbitrary and large images, scaling from tissue microarray (TMA) to whole slide image (WSI). Formally, SegGini constructs a tissue-graph representation for an input histology image, where the graph nodes depict tissue regions. Then, it performs weakly-supervised segmentation via node classification by using inexact image-level labels, incomplete scribbles, or both. We evaluated SegGini on two public prostate cancer datasets containing TMAs and WSIs. Our method achieved state-of-the-art segmentation performance on both datasets for various annotation settings while being comparable to a pathologist baseline., Comment: 10 pages, 5 figures

3D microscopy is key in the investigation of diverse biological systems, and the ever increasing availability of large datasets demands automatic cell identification methods that not only are accurate, but also can imply the uncertainty in their predictions to inform about potential errors and hence confidence in conclusions using them. While conventional deep learning methods often yield deterministic results, advances in deep Bayesian learning allow for accurate predictions with a probabilistic interpretation in numerous image classification and segmentation tasks. It is however nontrivial to extend such Bayesian methods to cell detection, which requires specialized learning frameworks. In particular, regression of density maps is a popular successful approach for extracting cell coordinates from local peaks in a postprocessing step, which hinders any meaningful probabilistic output. We herein propose a deep learning-based cell detection framework that can operate on large microscopy images and outputs desired probabilistic predictions by (i) integrating Bayesian techniques for the regression of uncertainty-aware density maps, where peak detection can be applied to generate cell proposals, and (ii) learning a mapping from the numerous proposals to a probabilistic space that is calibrated, i.e. accurately represents the chances of a successful prediction. Utilizing such calibrated predictions, we propose a probabilistic spatial analysis with Monte-Carlo sampling. We demonstrate this in revising an existing description of the distribution of a mesenchymal stromal cell type within the bone marrow, where our proposed methods allow us to reveal spatial patterns that are otherwise undetectable. Introducing such probabilistic analysis in quantitative microscopy pipelines will allow for reporting confidence intervals for testing biological hypotheses of spatial distributions.

Cancer diagnosis, prognosis, and therapy response predictions from tissue specimens highly depend on the phenotype and topological distribution of constituting histological entities. Thus, adequate tissue representations for encoding histological entities is imperative for computer aided cancer patient care. To this end, several approaches have leveraged cell-graphs that encode cell morphology and organization to denote the tissue information. These allow for utilizing machine learning to map tissue representations to tissue functionality to help quantify their relationship. Though cellular information is crucial, it is incomplete alone to comprehensively characterize complex tissue structure. We herein treat the tissue as a hierarchical composition of multiple types of histological entities from fine to coarse level, capturing multivariate tissue information at multiple levels. We propose a novel multi-level hierarchical entity-graph representation of tissue specimens to model hierarchical compositions that encode histological entities as well as their intra- and inter-entity level interactions. Subsequently, a graph neural network is proposed to operate on the hierarchical entity-graph representation to map the tissue structure to tissue functionality. Specifically, for input histology images we utilize well-defined cells and tissue regions to build HierArchical Cell-to-Tissue (HACT) graph representations, and devise HACT-Net, a graph neural network, to classify such HACT representations. As part of this work, we introduce the BReAst Carcinoma Subtyping (BRACS) dataset, a large cohort of H&E stained breast tumor images, to evaluate our proposed methodology against pathologists and state-of-the-art approaches. Through comparative assessment and ablation studies, our method is demonstrated to yield superior classification results compared to alternative methods as well as pathologists.

Fluorescence microscopy images contain several channels, each indicating a marker staining the sample. Since many different marker combinations are utilized in practice, it has been challenging to apply deep learning based segmentation models, which expect a predefined channel combination for all training samples as well as at inference for future application. Recent work circumvents this problem using a modality attention approach to be effective across any possible marker combination. However, for combinations that do not exist in a labeled training dataset, one cannot have any estimation of potential segmentation quality if that combination is encountered during inference. Without this, not only one lacks quality assurance but one also does not know where to put any additional imaging and labeling effort. We herein propose a method to estimate segmentation quality on unlabeled images by (i) estimating both aleatoric and epistemic uncertainties of convolutional neural networks for image segmentation, and (ii) training a Random Forest model for the interpretation of uncertainty features via regression to their corresponding segmentation metrics. Additionally, we demonstrate that including these uncertainty measures during training can provide an improvement on segmentation performance., Comment: Accepted at the IEEE International Symposium on Biomedical Imaging (ISBI) 2021. 4 pages and 4 figures

Purpose. Given the high level of expertise required for navigation and interpretation of ultrasound images, computational simulations can facilitate the training of such skills in virtual reality. With ray-tracing based simulations, realistic ultrasound images can be generated. However, due to computational constraints for interactivity, image quality typically needs to be compromised. Methods. We propose herein to bypass any rendering and simulation process at interactive time, by conducting such simulations during a non-time-critical offline stage and then learning image translation from cross-sectional model slices to such simulated frames. We use a generative adversarial framework with a dedicated generator architecture and input feeding scheme, which both substantially improve image quality without increase in network parameters. Integral attenuation maps derived from cross-sectional model slices, texture-friendly strided convolutions, providing stochastic noise and input maps to intermediate layers in order to preserve locality are all shown herein to greatly facilitate such translation task. Results. Given several quality metrics, the proposed method with only tissue maps as input is shown to provide comparable or superior results to a state-of-the-art that uses additional images of low-quality ultrasound renderings. An extensive ablation study shows the need and benefits from the individual contributions utilized in this work, based on qualitative examples and quantitative ultrasound similarity metrics. To that end, a local histogram statistics based error metric is proposed and demonstrated for visualization of local dissimilarities between ultrasound images.

Muscle transfers are commonly performed to restore joint function after muscle tears. However, there are not many biomechanical studies of muscle transfers, with the available ones often limited to passive movements in anatomical planes. Using data from three activities of daily living (ADL) and an available computational musculoskeletal model of the shoulder, we analyse the impact of a subscapularis tear and its treatment by a pectoralis major (PMA) transfer of the clavicular, sternal, or both segments. The shoulder model is validated against experimental data: the model kinematics with motion capture, muscle activity with EMG measurements, and model joint reaction force with in-vivo data from an instrumented prosthesis. Our results indicate that subscapularis tear requires a compensatory activation of the supraspinatus and is accompanied by a reduced co-contraction of the infraspinatus, both of which can be partially recovered after PMA transfer. Furthermore, although the PMA acts asynchronously to the subscapularis before the transfer, its patterns of activation change significantly after the transfer. Overall, this study demonstrates that muscle transfers have a significant impact on the neuromuscular system of the joint during active motion, beyond mere kinematics. Capability of a transferred muscle segment to activate similarly to intact subscapularis is found to be dependent on a considered motion. Differences in the activation patterns between intact subscapularis and segments of PMA may explain the difficulty of some patients in adapting their psycho-motor patterns during the rehabilitation period. Thereby, rehabilitation programs could benefit from targeted training on specific motion patterns and biofeedback programs. Finally, the condition of the anterior deltoid should be considered to avoid adding limitations to the joint function before a transfer of the clavicular part., Comment: 8 pages, 6 figures

Reverse Shoulder Arthroplasty (RSA) has become a very common procedure for shoulder joint replacement, even for scenarios where an anatomical reconstruction would traditionally be used. In this study, we investigate joint reaction forces and scapular kinematics for rotator cuff tears of different tendons with and without a reverse prosthesis. Available motion capture data during anterior flexion was input to a finite-element musculoskeletal shoulder model, and muscle activations were computed using inverse dynamics. The model was validated with respect to in-vivo glenohumeral joint reaction force (JRF) measurements, and also compared to existing clinical and biomechanical data. Simulations were carried out for the intact joint as well as for various tendons involved in a rotator cuff tear: superior (supraspinatus), superior-anterior (supraspinatus and subscapularis), and superior-posterior (supraspinatus, infraspinatus and teres minor). Each rotator cuff tear condition was repeated after shifting the humerus and the glenohumeral joint center of rotation to simulate the effect of a reverse prosthesis. Changes in compressive, shear, and total JRF were analysed, along with scapular upward rotation. The model compared favourably to in-vivo measurements, and existing clinical and biomechanical knowledge. Simulated JRF lie in the ranges of in-vivo JRF measurements and shows a linear increase past 90 degree flexion. Implanting a reverse prosthesis with a functional rotator cuff or with an isolated supraspinatus tear led to over 2 times higher compressive force component than with massive rotator cuff tears (superior-anterior or superior-posterior). Such higher compression forces might increase the risk of wear and implant fracture., Comment: 8 pages, 5 figures

Explainability of deep learning methods is imperative to facilitate their clinical adoption in digital pathology. However, popular deep learning methods and explainability techniques (explainers) based on pixel-wise processing disregard biological entities' notion, thus complicating comprehension by pathologists. In this work, we address this by adopting biological entity-based graph processing and graph explainers enabling explanations accessible to pathologists. In this context, a major challenge becomes to discern meaningful explainers, particularly in a standardized and quantifiable fashion. To this end, we propose herein a set of novel quantitative metrics based on statistics of class separability using pathologically measurable concepts to characterize graph explainers. We employ the proposed metrics to evaluate three types of graph explainers, namely the layer-wise relevance propagation, gradient-based saliency, and graph pruning approaches, to explain Cell-Graph representations for Breast Cancer Subtyping. The proposed metrics are also applicable in other domains by using domain-specific intuitive concepts. We validate the qualitative and quantitative findings on the BRACS dataset, a large cohort of breast cancer RoIs, by expert pathologists., Comment: CVPR 2021

Fluorescence microscopy allows for a detailed inspection of cells, cellular networks, and anatomical landmarks by staining with a variety of carefully-selected markers visualized as color channels. Quantitative characterization of structures in acquired images often relies on automatic image analysis methods. Despite the success of deep learning methods in other vision applications, their potential for fluorescence image analysis remains underexploited. One reason lies in the considerable workload required to train accurate models, which are normally specific for a given combination of markers, and therefore applicable to a very restricted number of experimental settings. We herein propose Marker Sampling and Excite, a neural network approach with a modality sampling strategy and a novel attention module that together enable (i) flexible training with heterogeneous datasets with combinations of markers and (ii) successful utility of learned models on arbitrary subsets of markers prospectively. We show that our single neural network solution performs comparably to an upper bound scenario where an ensemble of many networks is na\"ively trained for each possible marker combination separately. In addition, we demonstrate the feasibility of this framework in high-throughput biological analysis by revising a recent quantitative characterization of bone marrow vasculature in 3D confocal microscopy datasets and further confirm the validity of our approach on an additional, significantly different dataset of microvessels in fetal liver tissues. Not only can our work substantially ameliorate the use of deep learning in fluorescence microscopy analysis, but it can also be utilized in other fields with incomplete data acquisitions and missing modalities., Comment: Main: 21 pages, 6 figures, 1 table. Supplementary: 5 pages, 7 figures, 3 tables

To diagnose, plan, and treat musculoskeletal pathologies, understanding and reproducing muscle recruitment for complex movements is essential. With muscle activations for movements often being highly redundant, nonlinear, and time dependent, machine learning can provide a solution for their modeling and control for anatomy-specific musculoskeletal simulations. Sophisticated biomechanical simulations often require specialized computational environments, being numerically complex and slow, hindering their integration with typical deep learning frameworks. In this work, a deep reinforcement learning (DRL) based inverse dynamics controller is trained to control muscle activations of a biomechanical model of the human shoulder. In a generalizable end-to-end fashion, muscle activations are learned given current and desired position-velocity pairs. A customized reward functions for trajectory control is introduced, enabling straightforward extension to additional muscles and higher degrees of freedom. Using the biomechanical model, multiple episodes are simulated on a cluster simultaneously using the evolving neural models of the DRL being trained. Results are presented for a single-axis motion control of shoulder abduction for the task of following randomly generated angular trajectories.

Cancer diagnosis, prognosis, and therapeutic response prediction are heavily influenced by the relationship between the histopathological structures and the function of the tissue. Recent approaches acknowledging the structure-function relationship, have linked the structural and spatial patterns of cell organization in tissue via cell-graphs to tumor grades. Though cell organization is imperative, it is insufficient to entirely represent the histopathological structure. We propose a novel hierarchical cell-to-tissue-graph (HACT) representation to improve the structural depiction of the tissue. It consists of a low-level cell-graph, capturing cell morphology and interactions, a high-level tissue-graph, capturing morphology and spatial distribution of tissue parts, and cells-to-tissue hierarchies, encoding the relative spatial distribution of the cells with respect to the tissue distribution. Further, a hierarchical graph neural network (HACT-Net) is proposed to efficiently map the HACT representations to histopathological breast cancer subtypes. We assess the methodology on a large set of annotated tissue regions of interest from H\&E stained breast carcinoma whole-slides. Upon evaluation, the proposed method outperformed recent convolutional neural network and graph neural network approaches for breast cancer multi-class subtyping. The proposed entity-based topological analysis is more inline with the pathological diagnostic procedure of the tissue. It provides more command over the tissue modelling, therefore encourages the further inclusion of pathological priors into task-specific tissue representation.

Explainability of machine learning (ML) techniques in digital pathology (DP) is of great significance to facilitate their wide adoption in clinics. Recently, graph techniques encoding relevant biological entities have been employed to represent and assess DP images. Such paradigm shift from pixel-wise to entity-wise analysis provides more control over concept representation. In this paper, we introduce a post-hoc explainer to derive compact per-instance explanations emphasizing diagnostically important entities in the graph. Although we focus our analyses to cells and cellular interactions in breast cancer subtyping, the proposed explainer is generic enough to be extended to other topological representations in DP. Qualitative and quantitative analyses demonstrate the efficacy of the explainer in generating comprehensive and compact explanations., Comment: ICML'20 workshop on Computational Biology

We present a novel texture synthesis framework, enabling the generation of infinite, high-quality 3D textures given a 2D exemplar image. Inspired by recent advances in natural texture synthesis, we train deep neural models to generate textures by non-linearly combining learned noise frequencies. To achieve a highly realistic output conditioned on an exemplar patch, we propose a novel loss function that combines ideas from both style transfer and generative adversarial networks. In particular, we train the synthesis network to match the Gram matrices of deep features from a discriminator network. In addition, we propose two architectural concepts and an extrapolation strategy that significantly improve generalization performance. In particular, we inject both model input and condition into hidden network layers by learning to scale and bias hidden activations. Quantitative and qualitative evaluations on a diverse set of exemplars motivate our design decisions and show that our system performs superior to previous state of the art. Finally, we conduct a user study that confirms the benefits of our framework.

Speed-of-sound has been shown as a potential biomarker for breast cancer imaging, successfully differentiating malignant tumors from benign ones. Speed-of-sound images can be reconstructed from time-of-flight measurements from ultrasound images acquired using conventional handheld ultrasound transducers. Variational Networks (VN) have recently been shown to be a potential learning-based approach for optimizing inverse problems in image reconstruction. Despite earlier promising results, these methods however do not generalize well from simulated to acquired data, due to the domain shift. In this work, we present for the first time a VN solution for a pulse-echo SoS image reconstruction problem using diverging waves with conventional transducers and single-sided tissue access. This is made possible by incorporating simulations with varying complexity into training. We use loop unrolling of gradient descent with momentum, with an exponentially weighted loss of outputs at each unrolled iteration in order to regularize training. We learn norms as activation functions regularized to have smooth forms for robustness to input distribution variations. We evaluate reconstruction quality on ray-based and full-wave simulations as well as on tissue-mimicking phantom data, in comparison to a classical iterative (L-BFGS) optimization of this image reconstruction problem. We show that the proposed regularization techniques combined with multi-source domain training yield substantial improvements in the domain adaptation capabilities of VN, reducing median RMSE by 54% on a wave-based simulation dataset compared to the baseline VN. We also show that on data acquired from a tissue-mimicking breast phantom the proposed VN provides improved reconstruction in 12 milliseconds.

Ultrasound simulation based on ray tracing enables the synthesis of highly realistic images. It can provide an interactive environment for training sonographers as an educational tool. However, due to high computational demand, there is a trade-off between image quality and interactivity, potentially leading to sub-optimal results at interactive rates. In this work we introduce a deep learning approach based on adversarial training that mitigates this trade-off by improving the quality of simulated images with constant computation time. An image-to-image translation framework is utilized to translate low quality images into high quality versions. To incorporate anatomical information potentially lost in low quality images, we additionally provide segmentation maps to image translation. Furthermore, we propose to leverage information from acoustic attenuation maps to better preserve acoustic shadows and directional artifacts, an invaluable feature for ultrasound image interpretation. The proposed method yields an improvement of 7.2% in Fr\'{e}chet Inception Distance and 8.9% in patch-based Kullback-Leibler divergence.

Simulation-based ultrasound training can be an essential educational tool. Realistic ultrasound image appearance with typical speckle texture can be modeled as convolution of a point spread function with point scatterers representing tissue microstructure. Such scatterer distribution, however, is in general not known and its estimation for a given tissue type is fundamentally an ill-posed inverse problem. In this paper, we demonstrate a convolutional neural network approach for probabilistic scatterer estimation from observed ultrasound data. We herein propose to impose a known statistical distribution on scatterers and learn the mapping between ultrasound image and distribution parameter map by training a convolutional neural network on synthetic images. In comparison with several existing approaches, we demonstrate in numerical simulations and with in-vivo images that the synthesized images from scatterer representations estimated with our approach closely match the observations with varying acquisition parameters such as compression and rotation of the imaged domain.

Mitotic counting is a vital prognostic marker of tumor proliferation in breast cancer. Deep learning-based mitotic detection is on par with pathologists, but it requires large labeled data for training. We propose a deep classification framework for enhancing mitosis detection by leveraging class label information, via softmax loss, and spatial distribution information among samples, via distance metric learning. We also investigate strategies towards steadily providing informative samples to boost the learning. The efficacy of the proposed framework is established through evaluation on ICPR 2012 and AMIDA 2013 mitotic data. Our framework significantly improves the detection with small training data and achieves on par or superior performance compared to state-of-the-art methods for using the entire training data., Comment: 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI)

Attenuation of ultrasound waves varies with tissue composition, hence its estimation offers great potential for tissue characterization and diagnosis and staging of pathology. We recently proposed a method that allows to spatially reconstruct the distribution of the overall ultrasound attenuation in tissue based on computed tomography, using reflections from a passive acoustic reflector. This requires a standard ultrasound transducer operating in pulse-echo mode and a calibration protocol using water measurements, thus it can be implemented on conventional ultrasound systems with minor adaptations. Herein, we extend this method by additionally estimating and imaging the frequency-dependent nature of local ultrasound attenuation for the first time. Spatial distributions of attenuation coefficient and exponent are reconstructed, enabling an elaborate and expressive tissue-specific characterization. With simulations, we demonstrate that our proposed method yields a low reconstruction error of 0.04dB/cm at 1MHz for attenuation coefficient and 0.08 for the frequency exponent. With tissue-mimicking phantoms and ex-vivo bovine muscle samples, a high reconstruction contrast as well as reproducibility are demonstrated. Attenuation exponents of a gelatin-cellulose mixture and an ex-vivo bovine muscle sample were found to be, respectively, 1.4 and 0.5 on average, consistently from different images of their heterogeneous compositions. Such frequency-dependent parametrization could enable novel imaging and diagnostic techniques, as well as facilitate attenuation compensation of other ultrasound-based imaging techniques.

Shear-wave elastography (SWE) measures shear-wave speed (SWS), which is related to the underlying shear modulus of soft tissue. SWE methods generally assume that soft tissue viscoelasticity is independent of mechanical loading, however, soft tissues are known to have viscoelasticity changing nonlinearly with pre-compression. Hence, characterization by SWS alone is insufficient, where nonlinear properties can be seen as confounders but may also be utilized as additional bio-markers. The viscoelastic nature of a medium is fully characterized by its storage and loss moduli, which are related to SWS and shear-wave attenuation (SWA). In this work, we study SWA characteristics as a function of applied strain to measure nonlinear viscoelastic parameters in soft tissues. For this purpose, we apply incremental quasi-static compression on the samples while measuring SWS and SWA, from which we derive storage and loss moduli to estimate nonlinear viscoelastic parameters as a function of applied strain using acoustoelasticity (AE) theory. Results from tissue-mimicking phantoms with varying oil percentages and ex-vivo porcine liver experiments demonstrate the feasibility of the proposed approach. In both these experiments, SWA was observed to decrease with applied strain. For 10 % compression in ex-vivo livers, shear-wave attenuation decreased on average by 28 % (93 Np/m), while SWS increased on average by 20 % (0.26 m/s).

Bone surface delineation in ultrasound is of interest due to its potential in diagnosis, surgical planning, and post-operative follow-up in orthopedics, as well as the potential of using bones as anatomical landmarks in surgical navigation. We herein propose a method to encode the physics of ultrasound propagation into a factor graph formulation for the purpose of bone surface delineation. In this graph structure, unary node potentials encode the local likelihood for being a soft tissue or acoustic-shadow (behind bone surface) region, both learned through image descriptors. Pair-wise edge potentials encode ultrasound propagation constraints of bone surfaces given their large acoustic-impedance difference. We evaluate the proposed method in comparison with four earlier approaches, on in-vivo ultrasound images collected from dorsal and volar views of the forearm. The proposed method achieves an average root-mean-square error and symmetric Hausdorff distance of 0.28mm and 1.78mm, respectively. It detects 99.9% of the annotated bone surfaces with a mean scanline error (distance to annotations) of 0.39mm., Comment: 11 pages, 8 figures

Segmentation of anatomical structures is a fundamental image analysis task for many applications in the medical field. Deep learning methods have been shown to perform well, but for this purpose large numbers of manual annotations are needed in the first place, which necessitate prohibitive levels of resources that are often unavailable. In an active learning framework of selecting informed samples for manual labeling, expert clinician time for manual annotation can be optimally utilized, enabling the establishment of large labeled datasets for machine learning. In this paper, we propose a novel method that combines representativeness with uncertainty in order to estimate ideal samples to be annotated, iteratively from a given dataset. Our novel representativeness metric is based on Bayesian sampling, by using information-maximizing autoencoders. We conduct experiments on a shoulder magnetic resonance imaging (MRI) dataset for the segmentation of four musculoskeletal tissue classes. Quantitative results show that the annotation of representative samples selected by our proposed querying method yields an improved segmentation performance at each active learning iteration, compared to a baseline method that also employs uncertainty and representativeness metrics. For instance, with only 10% of the dataset annotated, our method reaches within 5% of Dice score expected from the upper bound scenario of all the dataset given as annotated (an impractical scenario due to resource constraints), and this gap drops down to a mere 2% when less than a fifth of the dataset samples are annotated. Such active learning approach to selecting samples to annotate enables an optimal use of the expert clinician time, being often the bottleneck in realizing machine learning solutions in medicine., Comment: 10 pages, 7 figures

Recent ultrasound imaging modalities based on ultrasound computed tomography indicate a huge potential to detect pathologies is tissue due to altered biomechanical properties. Especially the imaging of speed-of-sound (SoS) distribution in tissue has shown clinical promise and thus gained increasing attention in the field -- with several methods proposed based on transmission mode tomography. SoS imaging using conventional ultrasound (US) systems would be convenient and easy for clinical translation, but this requires using conventional US probes with single-sides tissue access and thus pulse-echo imaging sequences. Recent pulse-echo SoS imaging methods rely on plane wave (PW) insonifications, which is prone to strong aberration effects for non-homogeneous tissue composition. In this paper we propose to use diverging waves (DW) for SoS imaging and thus substantially improve the reconstruction of SoS distributions. We study this proposition by first plane wavefront aberrations compared to DW. We then present the sensitivity of both approaches to major parameterization choices on a set of simulated phantoms. Using the optimum parameter combination for each method for a given transducer model and imaging sequence, we analyze the SoS imaging performance comparatively between the two approaches. Results indicate that using DW instead of PW, the reconstruction accuracy improves substantially, by over 22% in reconstruction error (RMSE) and by 55% in contrast (CNR). We also demonstrate improvements in SoS reconstructions from an actual US acquisition of a breast phantom with tumor- and cyst-representative inclusions, with high and low SoS contrast, respectively.

For beamforming ultrasound (US) signals, typically a spatially constant speed-of-sound (SoS) is assumed to calculate delays. As SoS in tissue may vary relatively largely, this approximation may cause wavefront aberrations, thus degrading effective imaging resolution. In the literature, corrections have been proposed based on unidirectional SoS estimation or computationally-expensive a posteriori phase rectification. In this paper we demonstrate a direct delay correction approach for US beamforming, by leveraging 2D spatial SoS distribution estimates from plane-wave imaging. We show both in simulations and with ex vivo measurements that resolutions close to the wavelength limit can be achieved using our proposed local SoS-adaptive beamforming, yielding a lateral resolution improvement of 22% to 29% on tissue samples with up to 3% SoS-contrast (45m/s). We verify that our method accurately images absolute positions of tissue structures down to sub-pixel resolution of a tenth of a wavelength, whereas a global SoS assumption leads to artifactual localizations., Comment: will be published in the proceedings of the IEEE International Ultrasonics Symposium (IUS) 2019

Ultrasound attenuation is caused by absorption and scattering in tissue and is thus a function of tissue composition, hence its imaging offers great potential for screening and differential diagnosis. In this paper we propose a novel method that allows to reconstruct spatial attenuation distribution in tissue based on computed tomography, using reflections from a passive acoustic reflector. This requires a standard ultrasound transducer operating in pulse-echo mode, thus it can be implemented on conventional ultrasound systems with minor modifications. We use calibration with water measurements in order to normalize measurements for quantitative imaging of attenuation. In contrast to earlier techniques, we herein show that attenuation reconstructions are possible without any geometric prior on the inclusion location or shape. We present a quantitative evaluation of reconstructions based on simulations, gelatin phantoms, and ex-vivo bovine skeletal muscle tissue, achieving contrast-to-noise ratio of up to 2.3 for an inclusion in ex-vivo tissue., Comment: Accepted at MICCAI 2019 (International Conference on Medical Image Computing and Computer Assisted Intervention)

Tomographic image reconstruction is relevant for many medical imaging modalities including X-ray, ultrasound (US) computed tomography (CT) and photoacoustics, for which the access to full angular range tomographic projections might be not available in clinical practice due to physical or time constraints. Reconstruction from incomplete data in low signal-to-noise ratio regime is a challenging and ill-posed inverse problem that usually leads to unsatisfactory image quality. While informative image priors may be learned using generic deep neural network architectures, the artefacts caused by an ill-conditioned design matrix often have global spatial support and cannot be efficiently filtered out by means of convolutions. In this paper we propose to learn an inverse mapping in an end-to-end fashion via unrolling optimization iterations of a prototypical reconstruction algorithm. We herein introduce a network architecture that performs filtering jointly in both sinogram and spatial domains. To efficiently train such deep network we propose a novel regularization approach based on deep exponential weighting. Experiments on US and X-ray CT data show that our proposed method is qualitatively and quantitatively superior to conventional non-linear reconstruction methods as well as state-of-the-art deep networks for image reconstruction. Fast inference time of the proposed algorithm allows for sophisticated reconstructions in real-time critical settings, demonstrated with US SoS imaging of an ex vivo bovine phantom., Comment: Accepted to MICCAI 2019

In image processing tasks, spatial priors are essential for robust computations, regularization, algorithmic design and Bayesian inference. In this paper, we introduce weighted mean curvature (WMC) as a novel image prior and present an efficient computation scheme for its discretization in practical image processing applications. We first demonstrate the favorable properties of WMC, such as sampling invariance, scale invariance, and contrast invariance with Gaussian noise model; and we show the relation of WMC to area regularization. We further propose an efficient computation scheme for discretized WMC, which is demonstrated herein to process over 33.2 giga-pixels/second on GPU. This scheme yields itself to a convolutional neural network representation. Finally, WMC is evaluated on synthetic and real images, showing its superiority quantitatively to total-variation and mean curvature., Comment: 12 pages

Computational simulation of ultrasound (US) echography is essential for training sonographers. Realistic simulation of US interaction with microscopic tissue structures is often modeled by a tissue representation in the form of point scatterers, convolved with a spatially varying point spread function. This yields a realistic US B-mode speckle texture, given that a scatterer representation for a particular tissue type is readily available. This is often not the case and scatterers are nontrivial to determine. In this work we propose to estimate scatterer maps from sample US B-mode images of a tissue, by formulating this inverse mapping problem as image translation, where we learn the mapping with Generative Adversarial Networks, using a US simulation software for training. We demonstrate robust reconstruction results, invariant to US viewing and imaging settings such as imaging direction and center frequency. Our method is shown to generalize beyond the trained imaging settings, demonstrated on in-vivo US data. Our inference runs orders of magnitude faster than optimization-based techniques, enabling future extensions for reconstructing 3D B-mode volumes with only linear computational complexity.

Image-guided radiation therapy can benefit from accurate motion tracking by ultrasound imaging, in order to minimize treatment margins and radiate moving anatomical targets, e.g., due to breathing. One way to formulate this tracking problem is the automatic localization of given tracked anatomical landmarks throughout a temporal ultrasound sequence. For this, we herein propose a fully-convolutional Siamese network that learns the similarity between pairs of image regions containing the same landmark. Accordingly, it learns to localize and thus track arbitrary image features, not only predefined anatomical structures. We employ a temporal consistency model as a location prior, which we combine with the network-predicted location probability map to track a target iteratively in ultrasound sequences. We applied this method on the dataset of the Challenge on Liver Ultrasound Tracking (CLUST) with competitive results, where our work is the first to effectively apply CNNs on this tracking problem, thanks to our temporal regularization., Comment: Accepted at the IEEE International Symposium on Biomedical Imaging (ISBI) 2019

Comprehensive surgical planning require complex patient-specific anatomical models. For instance, functional muskuloskeletal simulations necessitate all relevant structures to be segmented, which could be performed in real-time using deep neural networks given sufficient annotated samples. Such large datasets of multiple structure annotations are costly to procure and are often unavailable in practice. Nevertheless, annotations from different studies and centers can be readily available, or become available in the future in an incremental fashion. We propose a class-incremental segmentation framework for extending a deep network trained for some anatomical structure to yet another structure using a small incremental annotation set. Through distilling knowledge from the current state of the framework, we bypass the need for a full retraining. This is a meta-method to extend any choice of desired deep segmentation network with only a minor addition per structure, which makes it suitable for lifelong class-incremental learning and applicable also for future deep neural network architectures. We evaluated our methods on a public knee dataset of 100 MR volumes. Through varying amount of incremental annotation ratios, we show how our proposed method can retain the previous anatomical structure segmentation performance superior to the conventional finetuning approach. In addition, our framework inherently exploits transferable knowledge from previously trained structures to incremental tasks, demonstrated by superior results compared to non-incremental training. With the presented method, new anatomical structures can be learned without catastrophic forgetting of older structures and without extensive increase of memory and complexity., Comment: Published in IJCARS. 8 pages, 4 figures, contains supplementary material