Abstract

Intra-retinal layer segmentation of Optical Coherence Tomography images is critical in the assessment of ocular diseases. Existing Energy minimization based methods employ handcrafted cost terms to define their energy and are not robust to the presence of abnormalities.We propose a novel, Linearly Parameterized, Conditional Random Field(LP-CRF) model whose energy is learnt from a set of training images in an end-to-end manner. The proposed LP-CRF comprises two convolution filter banks to capture the appearance of each tissue region and boundary, the relative weights of the shape priors and an additional term based on the appearance similarity of the adjacent boundary points. All the energy terms are jointly learnt using the Structured Support Vector Machine. The proposed method segments all retinal boundaries in a single step. Our method was evaluated on 107 Normal and 220 AMDB-scan images and found to outperform three publicly available OCT segmentation software. The average unsigned boundary localization error is 1.52±0.29 pixels for segmentation of 8 boundaries on the Normal data set and 1.9±0.65 pixels for 3 boundaries on the combined AMD and Normal data set establishing the robustness of the proposed method.

Abstract

Background and objective: Accurate segmentation of optic disc and cup from monocular color fundus images plays a significant role in the screening and diagnosis of glaucoma. Though optic cup is characterized by the drop in depth from the disc boundary, most existing methods segment the two structures separately and rely only on color and vessel kink based cues due to the lack of explicit depth information in color fundus images. Methods: We propose a novel boundary-based Conditional Random Field formulation that extracts both the optic disc and cup boundaries in a single optimization step. In addition to the color gradients, the proposed method explicitly models the depth which is estimated from the fundus image itself using a coupled, sparse dictionary trained on a set of image-depth map (derived from Optical Coherence Tomography) pairs. Results: The estimated depth achieved a correlation coefficient of 0.80 with respect to the ground truth. The proposed segmentation method outperformed several state-of-the-art methods on five public datasets. The average dice coefficient was in the range of 0.87–0.97 for disc segmentation across three datasets and 0.83 for cup segmentation on the DRISHTI-GS1 test set. The method achieved a good glaucoma classification performance with an average AUC of 0.85 for five fold cross-validation on RIM-ONE v2.

Abstract

Segmentation of retinal layers from Optical Coherence Tomography (OCT) volumes is a fundamental problem for any computer aided diagnostic algorithm development.This requires preprocessing steps such as denoising, region of interest extraction, flattening and edge detection all of which involve separate parameter tuning. In this paper,we explore deep learning techniques to automate all these steps and handle the presence/absence of pathologies. A model is proposed consisting of a combination of Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM). The CNN is used to extract layers of interest image and extract the edges, while the LSTM is used to trace the layer boundary. This model is trained on a mixture of normal and AMD cases using minimal data. Validation results on three public datasets show that the pixel-wise mean absolute error obtained with our system is1.30±0.48whichis lower than the inter-marker error of1.79±0.76. Our model’s performance is also on par with the existing methods

Abstract

Annotations are critical for machine learning and developing Computer Aided Detection (CAD) algorithms. However, a majority of medical data is either unlabeled or annotated only at the image-level. This poses a problem specifically for employing deep learning based approaches for CAD development as they require large amounts of annotated data for training. Data augmentation is a popular solution to address this need. We explore crowdsourcing as a solution for training a deep neural network (DNN) for lesion detection. Our solution employs a strategy to overcome the noisy nature of crowdsourced annotations by i) assigning a reliability factor for each subject of the crowd based on their performance (at global and local levels) and experience and ii) requiring region of interest(ROI) markings rather than pixel-level markings from the crowd. We present a solution for training the DNN with data drawn from a heterogeneous mixture of annotations, namely,very limited number of pixel-level markings by experts and crowdsourced ROI markings. Experimental results obtained for hard exudate detection from color fundus images show that training with processed/refined crowdsourced data is effective as detection performance improves by 25% over training with just expert-markings and by 11% over training with annotation derived using majority voting among the crowd

Abstract

SD-OCT is a non invasive cross sectional imaging modality useful for diagnosis of macular defects. Efficient detection and segmentation of the abnormalities seen as biomarkers in OCT can help in analyzing the progression of the disease and advising effective treatment for the associated disease. In this work we proposes a fully automated Generalized Motion Pattern(GMP) based segmentation method using a cascade of fully convolutional networks for detection and segmentation of retinal fluids from SD-OCT scans. General methods for segmentation depend on domain knowledge based feature extraction , whereas we pro-pose a method based on Generalized Motion Pattern (GMP) [1] which is derived by inducing motion to an image to suppress the background.The proposed method is parallelizable and handles inter-scanner variability efficiently. Our method achieves a mean Dice score of 0.61,0.70 and0.73 during segmentation and a mean AUC of 0.85,0.84 and 0.87 during detection for the 3 types of fluids IRF,SRF and PDE respectively

Abstract

Focus, depth-of-field, and defocus are important ele-ments that portray the aesthetic emphasis in a good pho-tograph. The ability to manipulate the focus after capture provides useful creative control to photographers. Captur-ing focal stacks - multiple images with small change in focus setting - of static scenes is relatively easy with modern cam-eras. We propose a compact representation for focal stacks using an all-in-focus image, a focal-slice index map and pair-wise defocus blur parameters. Using our representation, we show reconstruction of images with different focus effects including extended focus, multiple focus, and scene synthesis with natural focus effects. A user study shows high acceptability of the synthesized images compared to real ones. The compact and powerful representation of focal stacks makes them suitable for handling by image editing tools in order to provide flexible focus manipulation

Abstract

Inferring trust relations between social media users is critical for a number of applications wherein users seek credible information. The fact that available trust relations are scarce and skewed makes trust prediction a challenging task. To the best of our knowledge, this is the first work on exploring representation learning for trust prediction. We propose an approach that uses only a small amount of binary user-user trust relations to simultaneously learn user embeddings and a model to predict trust between user pairs. We empirically demonstrate that for trust prediction, our approach outperforms classifier-based approaches which use state-of-the-art representation learning methods like DeepWalk and LINE as features. We also conduct experiments which use embeddings pre-trained with DeepWalk and LINE each as an input to our model, resulting in further performance improvement. Experiments with a dataset of ~356K user pairs show that the proposed method can obtain a high F-score of 92.65%.

Abstract

Identifying medical persona from a social media post is of paramount importance for drug marketing and pharmacovigilance. In this work, we propose multiple approaches to infer the medical persona associated with a social media post. We pose this as a supervised multi-label text classification problem. The main challenge is to identify the hidden cues in a post that are indicative of a particular persona. We first propose a large set of manually engineered features for this task. Further, we propose multiple neural network based architectures to extract useful features from these posts using pre-trained word embeddings. Our experiments on thousands of blogs and tweets show that the proposed approach results in 7% and 5% gain in F-measure over manual feature engineering based approach for blogs and tweets respectively.

Abstract

Deep neural networks have yielded immense success in speech recognition, computer vision and natural language processing. However, the exploration of deep neural networks for recommender systems has received a relatively little introspection. Also, different recommendation scenarios have their own issues which creates the need for different approaches for recommendation. Specifically in news recommendation a major problem is that of varying user interests. In this work, we use deep neural networks with attention to tackle the problem of news recommendation. The key factor in user-item based collaborative filtering is to identify the interaction between user and item features. Matrix factorization is one of the most common approaches for identifying this interaction. It maps both the users and the items into a joint latent factor space such that user-item interactions in that space can be modeled as inner products in that space. Some recent work has used deep neural networks with the motive to learn an arbitrary function instead of the inner product that is used for capturing the user-item interaction. However, directly adapting it for the news domain does not seem to be very suitable. This is because of the dynamic nature of news readership where the interests of the users keep changing with time. Hence, it becomes challenging for recommendation systems to model both user preferences as well as account for the interests which keep changing over time. We present a deep neural model, where a non-linear mapping of users and item features are learnt first. For learning a non-linear mapping for the users we use an attention-based recurrent t layer in combination with fully connected layers. For learning the mappings for the items we use only fully connected layers. We then use a ranking based objective function to learn the parameters of the network. We also use the content of the news articles as features for our model. Extensive experiments on a real-world dataset show a significant improvement of our proposed model over the state-of-the-art by 4.7% (Hit Ratio@10). Along with this, we also show the effectiveness of our model to handle the user cold-start and item cold-start problems.

Abstract

Existing work for abstractive multidocument summarization utilise existing phrase structures directly extracted from input documents to generate summary sentences. These methods can suffer from lack of consistence and coherence in merging phrases. We introduce a novel approach for abstractive multidocument summarization through partial dependency tree extraction, recombination and linearization. The method entrusts the summarizer to generate its own topically coherent sequential structures from scratch for effective communication. Results on TAC 2011, DUC-2004 and 2005 show that our system gives competitive results compared with state of the art abstractive summarization approaches in the literature. We also achieve competitive results in linguistic quality assessed by human evaluators.