Regularized regression methods such as principal component or partial least squares regression perform well in learning tasks on high dimensional spectral data, but cannot explicitly eliminate ...irrelevant features. The random forest classifier with its associated Gini feature importance, on the other hand, allows for an explicit feature elimination, but may not be optimally adapted to spectral data due to the topology of its constituent classification trees which are based on orthogonal splits in feature space.
We propose to combine the best of both approaches, and evaluated the joint use of a feature selection based on a recursive feature elimination using the Gini importance of random forests' together with regularized classification methods on spectral data sets from medical diagnostics, chemotaxonomy, biomedical analytics, food science, and synthetically modified spectral data. Here, a feature selection using the Gini feature importance with a regularized classification by discriminant partial least squares regression performed as well as or better than a filtering according to different univariate statistical tests, or using regression coefficients in a backward feature elimination. It outperformed the direct application of the random forest classifier, or the direct application of the regularized classifiers on the full set of features.
The Gini importance of the random forest provided superior means for measuring feature relevance on spectral data, but - on an optimal subset of features - the regularized classifiers might be preferable over the random forest classifier, in spite of their limitation to model linear dependencies only. A feature selection based on Gini importance, however, may precede a regularized linear classification to identify this optimal subset of features, and to earn a double benefit of both dimensionality reduction and the elimination of noise from the classification task.
The landscapes of the Near East show both the first settlements and the longest trajectories of settlement systems. Mounding is a characteristic property of these settlement sites, resulting from ...millennia of continuing settlement activity at distinguished places. So far, however, this defining feature of ancient settlements has not received much attention, or even been the subject of systematic evaluation. We propose a remote sensing approach for comprehensively mapping the pattern of human settlement at large scale and establish the largest archaeological record for a landscape in Mesopotamia, mapping about 14,000 settlement sites—spanning eight millennia—at 15-m resolution in a 23,000-km2 area in northeastern Syria. To map both low- and high-mounded places—the latter of which are often referred to as "tells"—we develop a strategy for detecting anthrosols in time series of multispectral satellite images and measure the volume of settlement sites in a digital elevation model. Using this volume as a proxy to continued occupation, we find a dependency of the long-term attractiveness of a site on local water availability, but also a strong relation to the relevance within a basin-wide exchange network that we can infer from our record and third millennium B.C. intersite routes visible on the ground until recent times. We believe it is possible to establish a nearly comprehensive map of human settlements in the fluvial plains of northern Mesopotamia and beyond, and site volume may be a key quantity to uncover long-term trends in human settlement activity from such a record.
Whole-body imaging of mice is a key source of information for research. Organ segmentation is a prerequisite for quantitative analysis but is a tedious and error-prone task if done manually. Here, we ...present a deep learning solution called AIMOS that automatically segments major organs (brain, lungs, heart, liver, kidneys, spleen, bladder, stomach, intestine) and the skeleton in less than a second, orders of magnitude faster than prior algorithms. AIMOS matches or exceeds the segmentation quality of state-of-the-art approaches and of human experts. We exemplify direct applicability for biomedical research for localizing cancer metastases. Furthermore, we show that expert annotations are subject to human error and bias. As a consequence, we show that at least two independently created annotations are needed to assess model performance. Importantly, AIMOS addresses the issue of human bias by identifying the regions where humans are most likely to disagree, and thereby localizes and quantifies this uncertainty for improved downstream analysis. In summary, AIMOS is a powerful open-source tool to increase scalability, reduce bias, and foster reproducibility in many areas of biomedical research.
A new algorithm is presented for the automatic segmentation of Multiple Sclerosis (MS) lesions in 3D Magnetic Resonance (MR) images. It builds on a discriminative random decision forest framework to ...provide a voxel-wise probabilistic classification of the volume. The method uses multi-channel MR intensities (T1, T2, and FLAIR), knowledge on tissue classes and long-range spatial context to discriminate lesions from background. A symmetry feature is introduced accounting for the fact that some MS lesions tend to develop in an asymmetric way. Quantitative evaluation of the proposed methods is carried out on publicly available labeled cases from the MICCAI MS Lesion Segmentation Challenge 2008 dataset. When tested on the same data, the presented method compares favorably to all earlier methods. In an a posteriori analysis, we show how selected features during classification can be ranked according to their discriminative power and reveal the most important ones.
► Random forest for automatic segmentation of MS lesions in 3D MR images ► Features: multi-channel MR intensities, priors, long-range spatial context, symmetry ► Quantitative evaluation shows significant improvement over all earlier methods ► A ranking of the most discriminative features and channels is proposed ► The automatically learned decision sequence mimics the state-of-the-art pipeline
The zebrafish Danio rerio has become a popular model host to explore disease pathology caused by infectious agents. A main advantage is its transparency at an early age, which enables live imaging of ...infection dynamics. While multispecies infections are common in patients, the zebrafish model is rarely used to study them, although the model would be ideal for investigating pathogen-pathogen and pathogen-host interactions. This may be due to the absence of an established multispecies infection protocol for a defined organ and the lack of suitable image analysis pipelines for automated image processing. To address these issues, we developed a protocol for establishing and tracking single and multispecies bacterial infections in the inner ear structure (otic vesicle) of the zebrafish by imaging. Subsequently, we generated an image analysis pipeline that involved deep learning for the automated segmentation of the otic vesicle, and scripts for quantifying pathogen frequencies through fluorescence intensity measures. We used Pseudomonas aeruginosa, Acinetobacter baumannii, and Klebsiella pneumoniae, three of the difficult-to-treat ESKAPE pathogens, to show that our infection protocol and image analysis pipeline work both for single pathogens and pairwise pathogen combinations. Thus, our protocols provide a comprehensive toolbox for studying single and multispecies infections in real-time in zebrafish.
Despite great advances in brain tumor segmentation and clear clinical need, translation of state-of-the-art computational methods into clinical routine and scientific practice remains a major ...challenge. Several factors impede successful implementations, including data standardization and preprocessing. However, these steps are pivotal for the deployment of state-of-the-art image segmentation algorithms. To overcome these issues, we present BraTS Toolkit. BraTS Toolkit is a holistic approach to brain tumor segmentation and consists of three components: First, the BraTS Preprocessor facilitates data standardization and preprocessing for researchers and clinicians alike. It covers the entire image analysis workflow prior to tumor segmentation, from image conversion and registration to brain extraction. Second, BraTS Segmentor enables orchestration of BraTS brain tumor segmentation algorithms for generation of fully-automated segmentations. Finally, Brats Fusionator can combine the resulting candidate segmentations into consensus segmentations using fusion methods such as majority voting and iterative SIMPLE fusion. The capabilities of our tools are illustrated with a practical example to enable easy translation to clinical and scientific practice.
Variations in the shape and appearance of anatomical structures in medical images are often relevant radiological signs of disease. Automatic tools can help automate parts of this manual process. A ...cloud-based evaluation framework is presented in this paper including results of benchmarking current state-of-the-art medical imaging algorithms for anatomical structure segmentation and landmark detection: the VISCERAL Anatomy benchmarks. The algorithms are implemented in virtual machines in the cloud where participants can only access the training data and can be run privately by the benchmark administrators to objectively compare their performance in an unseen common test set. Overall, 120 computed tomography and magnetic resonance patient volumes were manually annotated to create a standard Gold Corpus containing a total of 1295 structures and 1760 landmarks. Ten participants contributed with automatic algorithms for the organ segmentation task, and three for the landmark localization task. Different algorithms obtained the best scores in the four available imaging modalities and for subsets of anatomical structures. The annotation framework, resulting data set, evaluation setup, results and performance analysis from the three VISCERAL Anatomy benchmarks are presented in this article. Both the VISCERAL data set and Silver Corpus generated with the fusion of the participant algorithms on a larger set of non-manually-annotated medical images are available to the research community.
Display omitted
► Automatic localization of landmarks in complex, repetitive anatomical structures. ► Random Forest classifiers for every landmark as a pre-filtering stage. ► Hough regression model ...for refining the landmark candidate positions. ► Parts-based model of global landmark topology to select the final landmark positions. ► Results on three challenging data sets, median residuals of 0.80mm, 1.19mm, 2.71mm.
The accurate localization of anatomical landmarks is a challenging task, often solved by domain specific approaches. We propose a method for the automatic localization of landmarks in complex, repetitive anatomical structures.
The key idea is to combine three steps: (1) a classifier for pre-filtering anatomical landmark positions that (2) are refined through a Hough regression model, together with (3) a parts-based model of the global landmark topology to select the final landmark positions. During training landmarks are annotated in a set of example volumes. A classifier learns local landmark appearance, and Hough regressors are trained to aggregate neighborhood information to a precise landmark coordinate position. A non-parametric geometric model encodes the spatial relationships between the landmarks and derives a topology which connects mutually predictive landmarks. During the global search we classify all voxels in the query volume, and perform regression-based agglomeration of landmark probabilities to highly accurate and specific candidate points at potential landmark locations. We encode the candidates’ weights together with the conformity of the connecting edges to the learnt geometric model in a Markov Random Field (MRF). By solving the corresponding discrete optimization problem, the most probable location for each model landmark is found in the query volume.
We show that this approach is able to consistently localize the model landmarks despite the complex and repetitive character of the anatomical structures on three challenging data sets (hand radiographs, hand CTs, and whole body CTs), with a median localization error of 0.80mm, 1.19mm and 2.71mm, respectively.
Display omitted
•Joint 3-D vessel segmentation and centerline extraction based on multivariate Hough voting and oblique random forests.•Steerable filters for efficient computation of local image ...features at different scales and orientations.•Novel oblique split model for effective parametrization of split weight sparsity superior to univariate orthogonal splits.•Systematic evaluation and validation on synthetic data and high-resolution imaging datasets of the rat visual cortex.•Learning-based approach proves highly accurate and robust in spite of high level of label noise in the training data.
We propose a novel framework for joint 3-D vessel segmentation and centerline extraction. The approach is based on multivariate Hough voting and oblique random forests (RFs) that we learn from noisy annotations. It relies on steerable filters for the efficient computation of local image features at different scales and orientations.
We validate both the segmentation performance and the centerline accuracy of our approach both on synthetic vascular data and four 3-D imaging datasets of the rat visual cortex at 700nm resolution. First, we evaluate the most important structural components of our approach: (1) Orthogonal subspace filtering in comparison to steerable filters that show, qualitatively, similarities to the eigenspace filters learned from local image patches. (2) Standard RF against oblique RF. Second, we compare the overall approach to different state-of-the-art methods for (1) vessel segmentation based on optimally oriented flux (OOF) and the eigenstructure of the Hessian, and (2) centerline extraction based on homotopic skeletonization and geodesic path tracing.
Our experiments reveal the benefit of steerable over eigenspace filters as well as the advantage of oblique split directions over univariate orthogonal splits. We further show that the learning-based approach outperforms different state-of-the-art methods and proves highly accurate and robust with regard to both vessel segmentation and centerline extraction in spite of the high level of label noise in the training data.
We hypothesized that machine learning analysis based on texture information from the preoperative MRI can predict IDH mutational status in newly diagnosed WHO grade II and III gliomas. This ...retrospective study included in total 79 consecutive patients with a newly diagnosed WHO grade II or III glioma. Local binary pattern texture features were generated from preoperative B0 and fractional anisotropy (FA) diffusion tensor imaging. Using a training set of 59 patients, a single hidden layer neural network was then trained on the texture features to predict IDH status. The model was validated based on the prediction accuracy calculated in a previously unseen set of 20 gliomas. Prediction accuracy of the generated model was 92% (54/59 cases; AUC = 0.921) in the training and 95% (19/20; AUC = 0.952) in the validation cohort. The ten most important features were comprised of tumor size and both B0 and FA texture information, underlining the joint contribution of imaging data to classification. Machine learning analysis of DTI texture information and tumor size reliably predicts IDH status in preoperative MRI of gliomas. Such information may increasingly support individualized surgical strategies, supplement pathological analysis and highlight the potential of radiogenomics.