Accurate cardiac motion estimation from cine cardiac magnetic resonance (CMR) images is vital for assessing cardiac function and detecting its abnormalities. Existing methods often struggle to accurately capture heart motion because they rely on intensity-based image registration similarity losses that may overlook cardiac anatomical regions. To address this, we propose CardioMorphNet, a recurrent Bayesian deep learning framework for 3D cardiac shape-guided deformable registration using short-axis (SAX) CMR images. It employs a recurrent variational autoencoder to model spatio-temporal dependencies across the cardiac cycle, along with two posterior models for bi-ventricular segmentation and motion estimation. The derived loss function from the Bayesian formulation guides the framework to focus on anatomical regions by recursively registering segmentation maps without using intensity-based image registration similarity loss, while leveraging sequential SAX volumes and spatio-temporal features. The Bayesian modelling also enables the computation of uncertainty maps for the estimated motion fields. Validated on the UK Biobank and M&M datasets by comparing warped mask shapes with ground-truth masks, CardioMorphNet demonstrates superior performance in cardiac motion estimation, outperforming state-of-the-art methods. Uncertainty assessment shows that it also yields lower uncertainty values for estimated motion fields in the cardiac region compared with other probabilistic-based cardiac registration methods, indicating higher confidence in its predictions. In addition, the clinical indices extraction assessment shows that CardioMorphNet estimates the clinical indices more accurately than other approaches.

General movement assessment (GMA) is a non-invasive tool for the early detection of brain dysfunction through the qualitative assessment of general movements, and the development of automated methods can broaden its application. However, mainstream pose-based automated GMA methods are prone to uncertainty due to limited high-quality data and noisy pose estimation, hindering clinical reliability without reliable uncertainty measures. In this work, we introduce UDF-GMA which explicitly models epistemic uncertainty in model parameters and aleatoric uncertainty from data noise for pose-based automated GMA. UDF-GMA effectively disentangles uncertainties by directly modelling aleatoric uncertainty and estimating epistemic uncertainty through Bayesian approximation. We further propose fusing these uncertainties with the embedded motion representation to enhance class separation. Extensive experiments on the Pmi-GMA benchmark dataset demonstrate the effectiveness and generalisability of the proposed approach in predicting poor repertoire.

We propose a novel recurrent variational netwo8=k]irk, SegMorph, to perform concurrent segmentation and motion estimation on cardiac cine magnetic resonance image (CMR) sequences. Our model establishes a recurrent latent space that captures spatiotemporal features from cine-MRI sequences for multitask inference and synthesis. The proposed model follows a recurrent variational auto-encoder framework and adopts a learnt prior from the temporal inputs. We utilise a multi-branch decoder to handle bi-ventricular segmentation and motion estimation simultaneously. In addition to the spatiotemporal features from the latent space, motion estimation enriches the supervision of sequential segmentation tasks by providing pseudo-ground truth. On the other hand, the segmentation branch helps with motion estimation by predicting deformation vector fields (DVFs) based on anatomical information. Experimental results demonstrate that the proposed method performs better than state-of-the-art approaches qualitatively and quantitatively for both segmentation and motion estimation tasks. We achieved an 81% average Dice Similarity Coefficient (DSC) and a less than 3.5 mm average Hausdorff distance on segmentation. Meanwhile, we achieved a motion estimation Dice Similarity Coefficient of over 79%, with approximately 0.14% of pixels displaying a negative Jacobian determinant in the estimated DVFs.

Generative modelling for shapes is a prerequisite for In-Silico Clinical Trials (ISCTs), which aim to cost-effectively validate medical device interventions using synthetic anatomical shapes, often represented as 3D surface meshes. However, constructing AI models to generate shapes closely resembling the real mesh samples is challenging due to variable vertex counts, connectivities, and the lack of dense vertex-wise correspondences across the training data. Employing graph representations for meshes, we develop a novel unsupervised geometric deep-learning model to establish refinable shape correspondences in a latent space, construct a population-derived atlas and generate realistic synthetic shapes. We additionally extend our proposed base model to a joint shape generative-clustering multi-atlas framework to incorporate further variability and preserve more details in the generated shapes. Experimental results using liver and left-ventricular models demonstrate the approach's applicability to computational medicine, highlighting its suitability for ISCTs through a comparative analysis.

Determining early-stage prognostic markers and stratifying patients for effective treatment are two key challenges for improving outcomes for melanoma patients. Previous studies have used tumour transcriptome data to stratify patients into immune subgroups, which were associated with differential melanoma specific survival and potential predictive biomarkers. However, acquiring transcriptome data is a time-consuming and costly process. Moreover, it is not routinely used in the current clinical workflow. Here, we attempt to overcome this by developing deep learning models to classify gigapixel haematoxylin and eosin (H&E) stained pathology slides, which are well established in clinical workflows, into these immune subgroups. We systematically assess six different multiple instance learning (MIL) frameworks, using five different image resolutions and three different feature extraction methods. We show that pathology-specific self-supervised models using 10x resolution patches generate superior representations for the classification of immune subtypes. In addition, in a primary melanoma dataset, we achieve a mean area under the receiver operating characteristic curve (AUC) of 0.80 for classifying histopathology images into 'high' or 'low immune' subgroups and a mean AUC of 0.82 in an independent TCGA melanoma dataset. Furthermore, we show that these models are able to stratify patients into 'high' and 'low immune' subgroups with significantly different melanoma specific survival outcomes (log rank test, P< 0.005). We anticipate that MIL methods will allow us to find new biomarkers of high importance, act as a tool for clinicians to infer the immune landscape of tumours and stratify patients, without needing to carry out additional expensive genetic tests.

Deformable image registration (DIR) can be used to track cardiac motion. Conventional DIR algorithms aim to establish a dense and non-linear correspondence between independent pairs of images. They are, nevertheless, computationally intensive and do not consider temporal dependencies to regulate the estimated motion in a cardiac cycle. In this paper, leveraging deep learning methods, we formulate a novel hierarchical probabilistic model, termed DragNet, for fast and reliable spatio-temporal registration in cine cardiac magnetic resonance (CMR) images and for generating synthetic heart motion sequences. DragNet is a variational inference framework, which takes an image from the sequence in combination with the hidden states of a recurrent neural network (RNN) as inputs to an inference network per time step. As part of this framework, we condition the prior probability of the latent variables on the hidden states of the RNN utilised to capture temporal dependencies. We further condition the posterior of the motion field on a latent variable from hierarchy and features from the moving image. Subsequently, the RNN updates the hidden state variables based on the feature maps of the fixed image and the latent variables. Different from traditional methods, DragNet performs registration on unseen sequences in a forward pass, which significantly expedites the registration process. Besides, DragNet enables generating a large number of realistic synthetic image sequences given only one frame, where the corresponding deformations are also retrieved. The probabilistic framework allows for computing spatio-temporal uncertainties in the estimated motion fields. Our results show that DragNet performance is comparable with state-of-the-art methods in terms of registration accuracy, with the advantage of offering analytical pixel-wise motion uncertainty estimation across a cardiac cycle and being a motion generator. We will make our code publicly available.

Modern radiotherapy delivers treatment plans optimised on an individual patient level, using CT-based 3D models of patient anatomy. This optimisation is fundamentally based on simple assumptions about the relationship between radiation dose delivered to the cancer (increased dose will increase cancer control) and normal tissue (increased dose will increase rate of side effects). The details of these relationships are still not well understood, especially for radiation-induced toxicity. We propose a convolutional neural network based on multiple instance learning to analyse toxicity relationships for patients receiving pelvic radiotherapy. A dataset comprising of 315 patients were included in this study; with 3D dose distributions, pre-treatment CT scans with annotated abdominal structures, and patient-reported toxicity scores provided for each participant. In addition, we propose a novel mechanism for segregating the attentions over space and dose/imaging features independently for a better understanding of the anatomical distribution of toxicity. Quantitative and qualitative experiments were performed to evaluate the network performance. The proposed network could predict toxicity with 80% accuracy. Attention analysis over space demonstrated that there was a significant association between radiation dose to the anterior and right iliac of the abdomen and patient-reported toxicity. Experimental results showed that the proposed network had outstanding performance for toxicity prediction, localisation and explanation with the ability of generalisation for an unseen dataset.

Generalized nucleus segmentation techniques can contribute greatly to reducing the time to develop and validate visual biomarkers for new digital pathology datasets. We summarize the results of MoNuSeg 2018 Challenge whose objective was to develop generalizable nuclei segmentation techniques in digital pathology. The challenge was an official satellite event of the MICCAI 2018 conference in which 32 teams with more than 80 participants from geographically diverse institutes participated. Contestants were given a training set with 30 images from seven organs with annotations of 21,623 individual nuclei. A test dataset with 14 images taken from seven organs, including two organs that did not appear in the training set was released without annotations. Entries were evaluated based on average aggregated Jaccard index (AJI) on the test set to prioritize accurate instance segmentation as opposed to mere semantic segmentation. More than half the teams that completed the challenge outperformed a previous baseline. Among the trends observed that contributed to increased accuracy were the use of color normalization as well as heavy data augmentation. Additionally, fully convolutional networks inspired by variants of U-Net, FCN, and Mask-RCNN were popularly used, typically based on ResNet or VGG base architectures. Watershed segmentation on predicted semantic segmentation maps was a popular post-processing strategy. Several of the top techniques compared favorably to an individual human annotator and can be used with confidence for nuclear morphometrics.

The recognition of different cell compartments, the types of cells, and their interactions is a critical aspect of quantitative cell biology. However, automating this problem has proven to be non-trivial and requires solving multi-class image segmentation tasks that are challenging owing to the high similarity of objects from different classes and irregularly shaped structures. To alleviate this, graphical models are useful due to their ability to make use of prior knowledge and model inter-class dependences. Directed acyclic graphs, such as trees, have been widely used to model top-down statistical dependences as a prior for improved image segmentation. However, using trees, a few inter-class constraints can be captured. To overcome this limitation, we propose polytree graphical models that capture label proximity relations more naturally compared to tree-based approaches. A novel recursive mechanism based on two-pass message passing was developed to efficiently calculate closed-form posteriors of graph nodes on polytrees. The algorithm is evaluated on simulated data and on two publicly available fluorescence microscopy datasets, outperforming directed trees and three state-of-the-art convolutional neural networks, namely, SegNet, DeepLab, and PSPNet. Polytrees are shown to outperform directed trees in predicting segmentation error by highlighting areas in the segmented image that do not comply with prior knowledge. This paves the way to uncertainty measures on the resulting segmentation and guide subsequent segmentation refinement.

Lesion segmentation is the first step in most automatic melanoma recognition systems. Deficiencies and difficulties in dermoscopic images such as color inconstancy, hair occlusion, dark corners, and color charts make lesion segmentation an intricate task. In order to detect the lesion in the presence of these problems, we propose a supervised saliency detection method tailored for dermoscopic images based on the discriminative regional feature integration (DRFI). A DRFI method incorporates multilevel segmentation, regional contrast, property, background descriptors, and a random forest regressor to create saliency scores for each region in the image. In our improved saliency detection method, mDRFI, we have added some new features to regional property descriptors. Also, in order to achieve more robust regional background descriptors, a thresholding algorithm is proposed to obtain a new pseudo-background region. Findings reveal that mDRFI is superior to DRFI in detecting the lesion as the salient object in dermoscopic images. The proposed overall lesion segmentation framework uses detected saliency map to construct an initial mask of the lesion through thresholding and postprocessing operations. The initial mask is then evolving in a level set framework to fit better on the lesion's boundaries. The results of evaluation tests on three public datasets show that our proposed segmentation method outperforms the other conventional state-of-the-art segmentation algorithms and its performance is comparable with most recent approaches that are based on deep convolutional neural networks.

A probabilistic group-wise similarity registration technique based on Student's t-mixture model (TMM) and a multi-resolution extension of the same (mr-TMM) are proposed in this study, to robustly align shapes and establish valid correspondences, for the purpose of training statistical shape models (SSMs). Shape analysis across large cohorts requires automatic generation of the requisite training sets. Automated segmentation and landmarking of medical images often result in shapes with varying proportions of outliers and consequently require a robust method of alignment and correspondence estimation. Both TMM and mrTMM are validated by comparison with state-of-the-art registration algorithms based on Gaussian mixture models (GMMs), using both synthetic and clinical data. Four clinical data sets are used for validation: (a) 2D femoral heads (K= 1000 samples generated from DXA images of healthy subjects); (b) control-hippocampi (K= 50 samples generated from T1-weighted magnetic resonance (MR) images of healthy subjects); (c) MCI-hippocampi (K= 28 samples generated from MR images of patients diagnosed with mild cognitive impairment); and (d) heart shapes comprising left and right ventricular endocardium and epicardium (K= 30 samples generated from short-axis MR images of: 10 healthy subjects, 10 patients diagnosed with pulmonary hypertension and 10 diagnosed with hypertrophic cardiomyopathy). The proposed methods significantly outperformed the state-of-the-art in terms of registration accuracy in the experiments involving synthetic data, with mrTMM offering significant improvement over TMM. With the clinical data, both methods performed comparably to the state-of-the-art for the hippocampi and heart data sets, which contained few outliers. They outperformed the state-of-the-art for the femur data set, containing large proportions of outliers, in terms of alignment accuracy, and the quality of SSMs trained, quantified in terms of generalization, compactness and specificity.

Inferring a probability density function (pdf) for shape from a population of point sets is a challenging problem. The lack of point-to-point correspondences and the non-linearity of the shape spaces undermine the linear models. Methods based on manifolds model the shape variations naturally, however, statistics are often limited to a single geodesic mean and an arbitrary number of variation modes. We relax the manifold assumption and consider a piece-wise linear form, implementing a mixture of distinctive shape classes. The pdf for point sets is defined hierarchically, modeling a mixture of Probabilistic Principal Component Analyzers (PPCA) in higher dimension. A Variational Bayesian approach is designed for unsupervised learning of the posteriors of point set labels, local variation modes, and point correspondences. By maximizing the model evidence, the numbers of clusters, modes of variations, and points on the mean models are automatically selected. Using the predictive distribution, we project a test shape to the spaces spanned by the local PPCA's. The method is applied to point sets from: i) synthetic data, ii) healthy versus pathological heart morphologies, and iii) lumbar vertebrae. The proposed method selects models with expected numbers of clusters and variation modes, achieving lower generalization-specificity errors compared to state-of-the-art.

Statistical shape modeling is a powerful tool for visualizing and quantifying geometric and functional patterns of the heart. After myocardial infarction (MI), the left ventricle typically remodels in response to physiological challenges. Several methods have been proposed in the literature to describe statistical shape changes. Which method best characterizes left ventricular remodeling after MI is an open research question. A better descriptor of remodeling is expected to provide a more accurate evaluation of disease status in MI patients. We therefore designed a challenge to test shape characterization in MI given a set of three-dimensional left ventricular surface points. The training set comprised 100 MI patients, and 100 asymptomatic volunteers (AV). The challenge was initiated in 2015 at the Statistical Atlases and Computational Models of the Heart workshop, in conjunction with the MICCAI conference. The training set with labels was provided to participants, who were asked to submit the likelihood of MI from a different (validation) set of 200 cases (100 AV and 100 MI). Sensitivity, specificity, accuracy and area under the receiver operating characteristic curve were used as the outcome measures. The goals of this challenge were to (1) establish a common dataset for evaluating statistical shape modeling algorithms in MI, and (2) test whether statistical shape modeling provides additional information characterizing MI patients over standard clinical measures. Eleven groups with a wide variety of classification and feature extraction approaches participated in this challenge. All methods achieved excellent classification results with accuracy ranges from 0.83 to 0.98. The areas under the receiver operating characteristic curves were all above 0.90. Four methods showed significantly higher performance than standard clinical measures. The dataset and software for evaluation are available from the Cardiac Atlas Project website1.

Three-dimensional reconstructions of coronary arteries can overcome some of the limitations of 2D X-ray angiography, namely artery overlap/foreshortening and lack of depth information. Model-based arterial reconstruction algorithms usually rely on 2D coronary artery segmentations and require good robustness to outliers. In this paper, we propose a novel probabilistic method to reconstruct coronary artery centrelines from retrospectively gated X-ray images based on a probabilistic mixture model. Specifically, 3D coronary artery centrelines are described by a mixture of Student’s t-distributions, and the reconstruction is formulated as maximum-likelihood estimation of the mixture model parameters, given the 2D segmentations of arteries from 2D X-ray images. Our method provides robustness against the erroneously segmented parts in the 2D segmentations by taking advantage of the inherent robustness of t-distributions. We validate our reconstruction results using synthetic phantom and clinical X-ray angiography data. The results show that the proposed method can cope with imperfect and noisy segmentation data.

A novel probabilistic, group-wise rigid registration framework is proposed in this study, to robustly align and establish correspondence across anatomical shapes represented as unstructured point sets. Student’s t-mixture model (TMM) is employed to exploit their inherent robustness to outliers. The primary application for such a framework is the automatic construction of statistical shape models (SSMs) of anatomical structures, from medical images. Tools used for automatic segmentation and landmarking of medical images often result in segmentations with varying proportions of outliers. The proposed approach is able to robustly align shapes and establish valid correspondences in the presence of considerable outliers and large variations in shape. A multi-resolution registration (mrTMM) framework is also formulated, to further improve the performance of the proposed TMM-based registration method. Comparisons with a state-of-the art approach using clinical data show that the mrTMM method in particular, achieves higher alignment accuracy and yields SSMs that generalise better to unseen shapes.

Computational fluid dynamics (CFD) is a valuable tool for studying vascular diseases, but requires long computational time. To alleviate this issue, we propose a statistical framework to predict the aneurysmal wall shear stress patterns directly from the aneurysm shape. A database of 38 complex intracranial aneurysm shapes is used to generate aneurysm morphologies and CFD simulations. The shapes and wall shear stresses are then converted to clouds of hybrid points containing both types of information. These are subsequently used to train a joint statistical model implementing a mixture of principal component analyzers. Given a new aneurysmal shape, the trained joint model is firstly collapsed to a shape only model and used to initialize the missing shear stress values. The estimated hybrid point set is further refined by projection to the joint model space. We demonstrate that our predicted patterns can achieve significant similarities to the CFD-based results. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Groupwise registration of point sets is the fundamental step in creating statistical shape models (SSMs). When the number of points on the sets varies across the population, each point set is often regarded as a spatially transformed Gaussian mixture model (GMM) sample, and the registration problem is formulated as the estimation of the underlying GMM from the training samples. Thus, each Gaussian in the mixture specifies a landmark (or model point), which is probabilistically corresponded to a training point. The Gaussian components, transformations, and probabilistic matches are often computed by an expectation-maximization (EM) algorithm. To avoid over- and under-fitting errors, the SSM should be optimized by tuning the required number of components. In this paper, rather than manually setting the number of components before training, we start from a maximal model and prune out the negligible points during the registration by a sparsity criterion. We show that by searching over the continuous space for optimal sparsity level, we can reduce the fitting errors (generalization and specificities), and thereby help the search process for a discrete number of model points. We propose an EM framework, adopting a symmetric Dirichlet distribution as a prior, to enforce sparsity on the mixture weights of Gaussians. The negligible model points are pruned by a quadratic programming technique during EM iterations. The proposed EM framework also iteratively updates the estimates of the rigid registration parameters of the point sets to the mean model. Next, we apply the principal component analysis to the registered and equal-length training point sets and construct the SSMs. This method is evaluated by learning of sparse SSMs from 15 manually segmented caudate nuclei, 24 hippocampal, and 20 prostate data sets. The generalization, specificity, and compactness of the proposed model favorably compare to a traditional EM based model.

We present a generative approach for simultaneously registering a probabilistic atlas of a healthy population to brain magnetic resonance (MR) scans showing glioma and segmenting the scans into tumor as well as healthy tissue labels. The proposed method is based on the expectation maximization (EM) algorithm that incorporates a glioma growth model for atlas seeding, a process which modifies the original atlas into one with tumor and edema adapted to best match a given set of patient's images. The modified atlas is registered into the patient space and utilized for estimating the posterior probabilities of various tissue labels. EM iteratively refines the estimates of the posterior probabilities of tissue labels, the deformation field and the tumor growth model parameters. Hence, in addition to segmentation, the proposed method results in atlas registration and a low-dimensional description of the patient scans through estimation of tumor model parameters. We validate the method by automatically segmenting 10 MR scans and comparing the results to those produced by clinical experts and two state-of-the-art methods. The resulting segmentations of tumor and edema outperform the results of the reference methods, and achieve a similar accuracy from a second human rater. We additionally apply the method to 122 patients scans and report the estimated tumor model parameters and their relations with segmentation and registration results. Based on the results from this patient population, we construct a statistical atlas of the glioma by inverting the estimated deformation fields to warp the tumor segmentations of patients scans into a common space.

This paper investigates the problem of atlas registration of brain images with gliomas. Multiparametric imaging modalities (T1, T1-CE, T2, and FLAIR) are first utilized for segmentations of different tissues, and to compute the posterior probability map (PBM) of membership to each tissue class, using supervised learning. Similar maps are generated in the initially normal atlas, by modeling the tumor growth, using reaction-diffusion equation. Deformable registration using a demons-like algorithm is used to register the patient images with the tumor bearing atlas. Joint estimation of the simulated tumor parameters (e.g., location, mass effect and degree of infiltration), and the spatial transformation is achieved by maximization of the log-likelihood of observation. An expectation-maximization algorithm is used in registration process to estimate the spatial transformation and other parameters related to tumor simulation are optimized through asynchronous parallel pattern search (APPSPACK). The proposed method has been evaluated on five simulated data sets created by statistically simulated deformations (SSD), and fifteen real multichannel glioma data sets. The performance has been evaluated both quantitatively and qualitatively, and the results have been compared to ORBIT, an alternative method solving a similar problem. The results show that our method outperforms ORBIT, and the warped templates have better similarity to patient images.

In this paper, a level-set-based geometric regularization method is proposed which has the ability to estimate the local orientation of the evolving front and utilize it as shape induced information for anisotropic propagation. We show that preserving anisotropic fronts can improve elongations of the extracted structures, while minimizing the risk of leakage. To that end, for an evolving front using its shape-offset level-set representation, a novel energy functional is defined. It is shown that constrained optimization of this functional results in an anisotropic expansion flow which is usefull for vessel segmentation. We have validated our method using synthetic data sets, 2-D retinal angiogram images and magnetic resonance angiography volumetric data sets. A comparison has been made with two state-of-the-art vessel segmentation methods. Quantitative results, as well as qualitative comparisons of segmentations, indicate that our regularization method is a promising tool to improve the efficiency of both techniques.