Search

The latest biological findings observe that the motionless "lock‐and‐key" theory is not generally applicable and that changes in atomic sites and binding pose can provide important information for understanding drug binding. However, the computational expenditure limits the growth of protein trajectory‐related studies, thus hindering the possibility of supervised learning. A spatial‐temporal pre‐training method based on the modified equivariant graph matching networks, dubbed ProtMD which has two specially designed self‐supervised learning tasks: atom‐level prompt‐based denoising generative task and conformation‐level snapshot ordering task to seize the flexibility information inside molecular dynamics (MD) trajectories with very fine temporal resolutions is presented. The ProtMD can grant the encoder network the capacity to capture the time‐dependent geometric mobility of conformations along MD trajectories. Two downstream tasks are chosen to verify the effectiveness of ProtMD through linear detection and task‐specific fine‐tuning. A huge improvement from current state‐of‐the‐art methods, with a decrease of 4.3% in root mean square error for the binding affinity problem and an average increase of 13.8% in the area under receiver operating characteristic curve and the area under the precision‐recall curve for the ligand efficacy problem is observed. The results demonstrate a strong correlation between the magnitude of conformation's motion in the 3D space and the strength with which the ligand binds with its receptor. [ABSTRACT FROM AUTHOR]

Detecting buildings, segmenting building footprints, and extracting building edges from high-resolution remote sensing images are vital in applications such as urban planning, change detection, smart cities, and map-making and updating. The tasks of building detection, footprint segmentation, and edge extraction affect each other to a certain extent. However, most previous works have focused on one of these three tasks and have lacked a multitask learning framework that can simultaneously solve the tasks of building detection, footprint segmentation and edge extraction, making it difficult to obtain smooth and complete buildings. This study proposes a novel multiscale and multitask deep learning framework to consider the dependencies among building detection, footprint segmentation, and edge extraction while completing all three tasks. In addition, a multitask feature fusion module is introduced into the deep learning framework to increase the robustness of feature extraction. A multitask loss function is also introduced to balance the training losses among the various tasks to obtain the best training results. Finally, the proposed method is applied to open-source building datasets and large-scale high-resolution remote sensing images and compared with other advanced building extraction methods. To verify the effectiveness of multitask learning, the performance of multitask learning and single-task training is compared in ablation experiments. The experimental results show that the proposed method has certain advantages over other methods and that multitask learning can effectively improve single-task performance. [ABSTRACT FROM AUTHOR]

As an essential role in cartographic generalization, road network selection produces basic geographic information across map scales. However, the previous selection methods could not simultaneously consider both attribute characteristics and spatial structure. In light of this, an intelligent road network selection method based on a graph neural network (GNN) is proposed in this paper. Firstly, the selection case is designed to construct a sample library. Secondly, some neighbor sampling and aggregation rules are developed to update road features. Then, a GNN-based selection model is designed to calculate classification labels, thus completing road network selection. Finally, a few comparative analyses with different selection methods are conducted, verifying that most of the accuracy values of the GNN model are stable over 90%. The experiments indicate that this method could aggregate stroke nodes and their neighbors together to synchronously preserve semantic, geometric, and topological features of road strokes, and the selection result is closer to the reference map. Therefore, this paper could bridge the distance between deep learning and cartographic generalization, thus facilitating a more intelligent road network selection method. [ABSTRACT FROM AUTHOR]

In modern neuroscience and clinical study, neuroscientists and clinicians often use non-invasive imaging techniques to validate theories and computational models, observe brain activities and diagnose brain disorders. The functional Magnetic Resonance Imaging (fMRI) is one of the commonly-used imaging modalities that can be used to understand human brain mechanisms as well as the diagnosis and treatment of brain disorders. The advances in artificial intelligence and the emergence of deep learning techniques have shown promising results to better interpret fMRI data. Deep learning techniques have rapidly become the state of the art for analyzing fMRI data sets and resulted in performance improvements in diverse fMRI applications. Deep learning is normally presented as an end-to-end learning process and can alleviate feature engineering requirements and hence reduce domain knowledge requirements to some extent. Under the framework of deep learning, fMRI data can be considered as images, time series or images series. Hence, different deep learning models such as convolutional neural networks, recurrent neural network, or a combination of both, can be developed to process fMRI data for different tasks. In this review, we discussed the basics of deep learning methods and focused on its successful implementations for brain disorder diagnosis based on fMRI images. The goal is to provide a high-level overview of brain disorder diagnosis with fMRI images from the perspective of deep learning applications. [ABSTRACT FROM AUTHOR]

• Comparative review on image preprocessing and attribute analysis techniques for malignant liver diagnosis. • Critical analysis of state-of-the-art liver CAD systems. • Collective overview of malignant liver CAD based on US, MRI and CT. The computer-based approach is required for the non-invasive detection of chronic liver diseases that are asymptomatic, progressive, and potentially fatal in nature. In this study, we review the computer-aided diagnosis of hepatic lesions in view of diffuse- and focal liver disorders. This survey mainly focuses on three image acquisition modalities: ultrasonography, computed tomography, and magnetic resonance imaging. We present the insightful analysis with pros and cons for each preliminary step, particularly preprocessing, attribute analysis, and classification techniques to accomplish clinical diagnostic tasks. In preprocessing, we explore and compare commonly used denoising, deblurring and segmentation methods. Denoising is mainly performed with nonlinear models. In contrast, deep neural networks are frequently applied for deblurring and automatic segmentation of region-of-interest. In attribute analysis, the most common approach comprises texture properties. For classification, the support vector machine is mainly utilized across three image acquisition modalities. However, comparative analysis shows the best performance is obtained by deep learning-based convolutional neural networks. Considering biopsy samples or pathological factors such as overall stage, margin, and differentiation can be helpful for improving the prediction performance. In addition, technique breakthrough is expected soon with advances in machine learning models to address data limitation problems and improve the prediction performance. [ABSTRACT FROM AUTHOR]

Autism spectrum disorder (ASD) is a neurological and developmental disorder. Traditional diagnosis of ASD is typically performed through the observation of behaviors and interview of a patient. However, these diagnosis methods are time-consuming and can be misleading sometimes. Integrating machine learning algorithms with neuroimages, a diagnosis method, can possibly be established to detect ASD subjects from typical control subjects. In this study, we develop deep learning methods for diagnosis of ASD from functional brain networks constructed with brain functional magnetic resonance imaging (fMRI) data. The entire Autism Brain Imaging Data Exchange 1 (ABIDE 1) data set is utilized to investigate the performance of our proposed methods. First, we construct the brain networks from brain fMRI images and define the raw features based on such brain networks. Second, we employ an autoencoder (AE) to learn the advanced features from the raw features. Third, we train a deep neural network (DNN) with the advanced features, which achieves the classification accuracy of 76.2% and the receiving operating characteristic curve (AUC) of 79.7%. As a comparison, we also apply the same advanced features to train several traditional machine learning algorithms to benchmark the classification performance. Finally, we combine the DNN with the pretrained AE and train it with the raw features, which achieves the classification accuracy of 79.2% and the AUC of 82.4%. These results show that our proposed deep learning methods outperform the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Motivation Protein–protein interactions (PPIs) play important roles in many biological processes. Conventional biological experiments for identifying PPI sites are costly and time-consuming. Thus, many computational approaches have been proposed to predict PPI sites. Existing computational methods usually use local contextual features to predict PPI sites. Actually, global features of protein sequences are critical for PPI site prediction. Results A new end-to-end deep learning framework, named DeepPPISP, through combining local contextual and global sequence features, is proposed for PPI site prediction. For local contextual features, we use a sliding window to capture features of neighbors of a target amino acid as in previous studies. For global sequence features, a text convolutional neural network is applied to extract features from the whole protein sequence. Then the local contextual and global sequence features are combined to predict PPI sites. By integrating local contextual and global sequence features, DeepPPISP achieves the state-of-the-art performance, which is better than the other competing methods. In order to investigate if global sequence features are helpful in our deep learning model, we remove or change some components in DeepPPISP. Detailed analyses show that global sequence features play important roles in DeepPPISP. Availability and implementation The DeepPPISP web server is available at http://bioinformatics.csu.edu.cn/PPISP/. The source code can be obtained from https://github.com/CSUBioGroup/DeepPPISP. Supplementary information Supplementary data are available at Bioinformatics online. [ABSTRACT FROM AUTHOR]

Motivation Many studies have shown that microRNAs (miRNAs) play a key role in human diseases. Meanwhile, traditional experimental methods for miRNA–disease association identification are extremely costly, time-consuming and challenging. Therefore, many computational methods have been developed to predict potential associations between miRNAs and diseases. However, those methods mainly predict the existence of miRNA–disease associations, and they cannot predict the deep-level miRNA–disease association types. Results In this study, we propose a new end-to-end deep learning method (called PDMDA) to predict deep-level miRNA–disease associations with graph neural networks (GNNs) and miRNA sequence features. Based on the sequence and structural features of miRNAs, PDMDA extracts the miRNA feature representations by a fully connected network (FCN). The disease feature representations are extracted from the disease–gene network and gene–gene interaction network by GNN model. Finally, a multilayer with three fully connected layers and a softmax layer is designed to predict the final miRNA–disease association scores based on the concatenated feature representations of miRNAs and diseases. Note that PDMDA does not take the miRNA–disease association matrix as input to compute the Gaussian interaction profile similarity. We conduct three experiments based on six association type samples (including circulations, epigenetics, target, genetics, known association of which their types are unknown and unknown association samples). We conduct fivefold cross-validation validation to assess the prediction performance of PDMDA. The area under the receiver operating characteristic curve scores is used as metric. The experiment results show that PDMDA can accurately predict the deep-level miRNA–disease associations. Availability and implementation Data and source codes are available at https://github.com/27167199/PDMDA. [ABSTRACT FROM AUTHOR]

Precise segmentation is in demand for hepatocellular carcinoma or metastasis clinical diagnosis due to the heterogeneous appearance and diverse anatomy of the liver on scanned abdominal computed tomography (CT) images. In this study, we present an automatic unified registration-free deep-learning-based model with residual block and dilated convolution for training end-to-end liver and lesion segmentation. A multi-scale approach has also been utilized to explore novel inter-slice features with multi-channel input images. A novel objective function is introduced to deal with fore- and background pixels imbalance based on the joint metric of dice coefficient and absolute volumetric difference. Further, batch normalization is used to improve the learning without any loss of useful information. The proposed methodology is extensively validated and tested on 30% of the publicly available Dircadb, LiTS, Sliver07, and Chaos datasets. A comparative analysis is conducted based on multiple evaluation metrics frequently used in segmentation competitions. The results show substantial improvement, with mean dice scores of 97.31, 97.38, 97.39 and 95.49% for the Dircadb, LiTS, Sliver07, and Chaos liver test sets, and 91.92 and 86.70% for Dircadb and LiTS lesion segmentation. It should be noted that we achieve the best lesion segmentation performance on common datasets. The obtained qualitative and quantitative results demonstrate that our proposed model outperform other state-of-the-art methods for liver and lesion segmentation, with competitive performance on additional datasets. Henceforth, it is envisaged as being applicable to pertinent medical segmentation applications. [ABSTRACT FROM AUTHOR]

Automatic International Classification of Diseases (ICD) coding is an important task in the future of artificial intelligence healthcare. In recent years, a lot of traditional machine learning-based methods have been proposed, and they achieved good results on this task. However, these traditional machine learning-based methods for automatic ICD coding only focus on the semantic features of clinical notes and ignore the feature extraction of ICD titles that are the descriptions of ICD codes. In this paper, we propose a knowledge attention-based deep learning framework called KAICD for automatic ICD coding. KAICD makes full use of the clinic notes and the ICD titles. The semantic features of clinic notes are extracted by a multi-scale convolutional neural network. For ICD titles, we use attention-based Bidirectional Gated Recurrent Unit (Bi-GRU) to build a knowledge database, which can offer additional information. Depending on input clinic notes, we can use the attention mechanism to obtain different knowledge vectors from the knowledge database where some ICD titles are more relevant to the input clinic notes. Last, we concatenate the knowledge vectors and the semantic features of clinic notes, and use them for the final prediction. KAICD is tested on a public dataset Medical Information Mart for Intensive Care III (MIMIC III); it achieves micro-precision of 0.502, micro-recall of 0.428, and micro-f1 of 0.462, which outperforms other competing methods. Furthermore, the results of the ablation study show that the knowledge database of ICD titles learned by the attention-based Bi-GRU enhances the feature expression and improves the prediction performance. [ABSTRACT FROM AUTHOR]

Long non-coding RNAs (lncRNAs) are a class of RNA molecules with more than 200 nucleotides. A growing amount of evidence reveals that subcellular localization of lncRNAs can provide valuable insights into their biological functions. Existing computational methods for predicting lncRNA subcellular localization use k -mer features to encode lncRNA sequences. However, the sequence order information is lost by using only k -mer features. We proposed a deep learning framework, DeepLncLoc, to predict lncRNA subcellular localization. In DeepLncLoc, we introduced a new subsequence embedding method that keeps the order information of lncRNA sequences. The subsequence embedding method first divides a sequence into some consecutive subsequences and then extracts the patterns of each subsequence, last combines these patterns to obtain a complete representation of the lncRNA sequence. After that, a text convolutional neural network is employed to learn high-level features and perform the prediction task. Compared with traditional machine learning models, popular representation methods and existing predictors, DeepLncLoc achieved better performance, which shows that DeepLncLoc could effectively predict lncRNA subcellular localization. Our study not only presented a novel computational model for predicting lncRNA subcellular localization but also introduced a new subsequence embedding method which is expected to be applied in other sequence-based prediction tasks. The DeepLncLoc web server is freely accessible at http://bioinformatics.csu.edu.cn/DeepLncLoc/ , and source code and datasets can be downloaded from https://github.com/CSUBioGroup/DeepLncLoc. [ABSTRACT FROM AUTHOR]

• Variational graph auto-encoders are excellent for predicting miRNA-disease associations. • Graph convolutional networks obtain good representations for miRNAs and diseases. • Variational auto-encoders can deal with missing data in the miRNA-disease network. • Integrating different databases helps predict novel miRNA-disease associations. Cumulative experimental studies have demonstrated the critical roles of microRNAs (miRNAs) in the diverse fundamental and important biological processes, and in the development of numerous complex human diseases. Thus, exploring the relationships between miRNAs and diseases is helpful with understanding the mechanisms, the detection, diagnosis, and treatment of complex diseases. As the identification of miRNA-disease associations via traditional biological experiments is time-consuming and expensive, an effective computational prediction method is appealing. In this study, we present a deep learning framework with variational graph auto-encoder for miRNA-disease association prediction (VGAE-MDA). VGAE-MDA first gets the representations of miRNAs and diseases from the heterogeneous networks constructed by miRNA-miRNA similarity, disease-disease similarity, and known miRNA-disease associations. Then, VGAE-MDA constructs two sub-networks: miRNA-based network and disease-based network. Combining the representations based on the heterogeneous network, two variational graph auto-encoders (VGAE) are deployed for calculating the miRNA-disease association scores from two sub-networks, respectively. Lastly, VGAE-MDA obtains the final predicted association score for a miRNA-disease pair by integrating the scores from these two trained networks. Unlike the previous model, the VGAE-MDA can mitigate the effect of noises from random selection of negative samples. Besides, the use of graph convolutional neural (GCN) network can naturally incorporate the node features from the graph structure while the variational autoencoder (VAE) makes use of latent variables to predict associations from the perspective of data distribution. The experimental results show that VGAE-MDA outperforms the state-of-the-art approaches in miRNA-disease association prediction. Besides, the effectiveness of our model has been further demonstrated by case studies. [ABSTRACT FROM AUTHOR]

Biomolecules, such as microRNAs, circRNAs, lncRNAs and genes, are functionally interdependent in human cells, and all play critical roles in diverse fundamental and vital biological processes. The dysregulations of such biomolecules can cause diseases. Identifying the associations between biomolecules and diseases can uncover the mechanisms of complex diseases, which is conducive to their diagnosis, treatment, prognosis and prevention. Due to the time consumption and cost of biologically experimental methods, many computational association prediction methods have been proposed in the past few years. In this study, we provide a comprehensive review of machine learning-based approaches for predicting disease–biomolecule associations with multi-view data sources. Firstly, we introduce some databases and general strategies for integrating multi-view data sources in the prediction models. Then we discuss several feature representation methods for machine learning-based prediction models. Thirdly, we comprehensively review machine learning-based prediction approaches in three categories: basic machine learning methods, matrix completion-based methods and deep learning-based methods, while discussing their advantages and disadvantages. Finally, we provide some perspectives for further improving biomolecule–disease prediction methods. [ABSTRACT FROM AUTHOR]

Drug repositioning can drastically decrease the cost and duration taken by traditional drug research and development while avoiding the occurrence of unforeseen adverse events. With the rapid advancement of high-throughput technologies and the explosion of various biological data and medical data, computational drug repositioning methods have been appealing and powerful techniques to systematically identify potential drug-target interactions and drug-disease interactions. In this review, we first summarize the available biomedical data and public databases related to drugs, diseases and targets. Then, we discuss existing drug repositioning approaches and group them based on their underlying computational models consisting of classical machine learning, network propagation, matrix factorization and completion, and deep learning based models. We also comprehensively analyze common standard data sets and evaluation metrics used in drug repositioning, and give a brief comparison of various prediction methods on the gold standard data sets. Finally, we conclude our review with a brief discussion on challenges in computational drug repositioning, which includes the problem of reducing the noise and incompleteness of biomedical data, the ensemble of various computation drug repositioning methods, the importance of designing reliable negative samples selection methods, new techniques dealing with the data sparseness problem, the construction of large-scale and comprehensive benchmark data sets and the analysis and explanation of the underlying mechanisms of predicted interactions. [ABSTRACT FROM AUTHOR]

• A novel hybrid computational framework for lncRNA-disease association prediction. • Improved method that combines matrix factorization and deep learning techniques. • Show good results in term of AUC and AUPR compared with the other methods. In recent years, accumulating studies have shown that long non-coding RNAs (lncRNAs) not only play an important role in the regulation of various biological processes but also are the foundation for understanding mechanisms of human diseases. Due to the high cost of traditional biological experiments, the number of experimentally verified lncRNA-disease associations is very limited. Thus, many computational approaches have been proposed to discover the underlying associations between lncRNAs and diseases. However, the associations between lncRNAs and diseases are too complicated to model by using only traditional matrix factorization-based methods. In this study, we propose a hybrid computational framework (SDLDA) for the lncRNA-disease association prediction. In our computational framework, we use singular value decomposition and deep learning to extract linear and non-linear features of lncRNAs and diseases, respectively. Then we train SDLDA by combing the linear and non-linear features. Compared to previous computational methods, the combination of linear and non-linear features reinforces each other, which is better than using only either matrix factorization or deep learning. The computational results show that SDLDA has a better performance over existing methods in the leave-one-out cross-validation. Furthermore, the case studies show that 28 out of 30 cancer-related lncRNAs (10 for gastric cancer, 10 for colon cancer and 8 for renal cancer) are verified by mining recent biomedical literature. Code and data can be accessed at https://github.com/CSUBioGroup/SDLDA. [ABSTRACT FROM AUTHOR]

Motivation Computationally predicting disease genes helps scientists optimize the in-depth experimental validation and accelerates the identification of real disease-associated genes. Modern high-throughput technologies have generated a vast amount of omics data, and integrating them is expected to improve the accuracy of computational prediction. As an integrative model, multimodal deep belief net (DBN) can capture cross-modality features from heterogeneous datasets to model a complex system. Studies have shown its power in image classification and tumor subtype prediction. However, multimodal DBN has not been used in predicting disease–gene associations. Results In this study, we propose a method to predict disease–gene associations by multimodal DBN (dgMDL). Specifically, latent representations of protein-protein interaction networks and gene ontology terms are first learned by two DBNs independently. Then, a joint DBN is used to learn cross-modality representations from the two sub-models by taking the concatenation of their obtained latent representations as the multimodal input. Finally, disease–gene associations are predicted with the learned cross-modality representations. The proposed method is compared with two state-of-the-art algorithms in terms of 5-fold cross-validation on a set of curated disease–gene associations. dgMDL achieves an AUC of 0.969 which is superior to the competing algorithms. Further analysis of the top-10 unknown disease–gene pairs also demonstrates the ability of dgMDL in predicting new disease–gene associations. Availability and implementation Prediction results and a reference implementation of dgMDL in Python is available on https://github.com/luoping1004/dgMDL. Supplementary information Supplementary data are available at Bioinformatics online. [ABSTRACT FROM AUTHOR]

Circular RNAs (circRNAs) are extensively expressed in cells and tissues, and play crucial roles in human diseases and biological processes. Recent studies have reported that circRNAs could function as RNA binding protein (RBP) sponges, meanwhile RBPs can also be involved in back-splicing. The interaction with RBPs is also considered an important factor for investigating the function of circRNAs. Hence, it is necessary to understand the interaction mechanisms of circRNAs and RBPs, especially in human cancers. Here, we present a novel method based on deep learning to identify cancer-specific circRNA–RBP binding sites (CSCRSites), only using the nucleotide sequences as the input. In CSCRSites, an architecture with multiple convolution layers is utilized to detect the features of the raw circRNA sequence fragments, and further identify the binding sites through a fully connected layer with the softmax output. The experimental results show that CSCRSites outperform the conventional machine learning classifiers and some representative deep learning methods on the benchmark data. In addition, the features learnt by CSCRSites are converted to sequence motifs, some of which can match to human known RNA motifs involved in human diseases, especially cancer. Therefore, as a deep learning-based tool, CSCRSites could significantly contribute to the function analysis of cancer-associated circRNAs. [ABSTRACT FROM AUTHOR]

Clinical named entity recognition (CNER) is a fundamental step for many clinical Natural Language Processing (NLP) systems, which aims to recognize and classify clinical entities such as diseases, symptoms, exams, body parts and treatments in clinical free texts. In recent years, with the development of deep learning technology, deep neural networks (DNNs) have been widely used in Chinese clinical named entity recognition and many other clinical NLP tasks. However, these state-of-the-art models failed to make full use of the global information and multi-level semantic features in clinical texts. We design an improved character-level representation approach which integrates the character embedding and the character-label embedding to enhance the specificity and diversity of feature representations. Then, a multi-head self-attention based Bi-directional Long Short-Term Memory Conditional Random Field (MUSA-BiLSTM-CRF) model is proposed. By introducing the multi-head self-attention and combining a medical dictionary, the model can more effectively capture the weight relationships between characters and multi-level semantic feature information, which is expected to greatly improve the performance of Chinese clinical named entity recognition. We evaluate our model on two CCKS challenge (CCKS2017 Task 2 and CCKS2018 Task 1) benchmark datasets and the experimental results show that our proposed model achieves the best performance competing with the state-of-the-art DNN based methods. [ABSTRACT FROM AUTHOR]

Although training a deep network with financial time series is not hard, the important issue is, how much the prediction for the truly new data can be trusted with a trained network. In this study, we propose a dynamic predictor selection algorithm (DPSA) that dynamically evaluates and selects the prediction model (predictor) for stock daily movement trend prediction. We first build an initial set of potential candidate predictors based on the convolutional long short-term memory networks (ConvLSTMs) by using different values of parameters. To evaluate the candidate predictors, we propose a kernel time-weighted fuzzy c-means clustering algorithm (KTFCM), which improves the kernel FCM algorithm (KFCM), to organize the historical samples according to their relevance to the target sample, which makes the historical samples that are closely related to the target sample have more influence on the predictors. Then, we use the well-organized historical samples to evaluate the candidate predictors. The predictor that yields the best accuracy is selected to predict the target sample. The proposed DPSA algorithm takes less than one minute in total for training the networks, evaluating and selecting the predictors, and performing prediction, which greatly shortens the time of the deep learning prediction. We perform the comparative experiments for the proposed DPSA algorithm and seven popular methods. These experiments test a large real-life financial time series data of various stock markets. The experiment results show that DPSA achieves the best accuracy and the highest return compared to the seven other popular methods. • A dynamic predictor selection model for financial prediction is proposed. • The model organizes the historical samples according to their relevance. • A clustering algorithm is proposed to improve prediction performance. • The model needs a few iterations to differentiate each of the predictor. • The training time of this model is greatly shorter than in deep learning methods. [ABSTRACT FROM AUTHOR]

• Mammographic breast density is a well-established breast cancer risk factor. • In a large multi-ethnic population from two different sites, Deep-LIBRA demonstrated high accuracy in estimating breast density. • It improved density-based discrimination of breast cancer over state-of-the art research and commercial methods. • Deep-LIBRA provides researchers across the world a reliable, reproducible and robust breast density measure with spatial segmentation maps for various studies. • Deep-LIBRA is an important advancement towards personalized breast cancer risk calculations in the clinical setting. [Display omitted] Breast density is an important risk factor for breast cancer that also affects the specificity and sensitivity of screening mammography. Current federal legislation mandates reporting of breast density for all women undergoing breast cancer screening. Clinically, breast density is assessed visually using the American College of Radiology Breast Imaging Reporting And Data System (BI-RADS) scale. Here, we introduce an artificial intelligence (AI) method to estimate breast density from digital mammograms. Our method leverages deep learning using two convolutional neural network architectures to accurately segment the breast area. An AI algorithm combining superpixel generation and radiomic machine learning is then applied to differentiate dense from non-dense tissue regions within the breast, from which breast density is estimated. Our method was trained and validated on a multi-racial, multi-institutional dataset of 15,661 images (4,437 women), and then tested on an independent matched case-control dataset of 6368 digital mammograms (414 cases; 1178 controls) for both breast density estimation and case-control discrimination. On the independent dataset, breast percent density (PD) estimates from Deep-LIBRA and an expert reader were strongly correlated (Spearman correlation coefficient = 0.90). Moreover, in a model adjusted for age and BMI, Deep-LIBRA yielded a higher case-control discrimination performance (area under the ROC curve, AUC = 0.612 [95% confidence interval (CI): 0.584, 0.640]) compared to four other widely-used research and commercial breast density assessment methods (AUCs = 0.528 to 0.599). Our results suggest a strong agreement of breast density estimates between Deep-LIBRA and gold-standard assessment by an expert reader, as well as improved performance in breast cancer risk assessment over state-of-the-art open-source and commercial methods. [ABSTRACT FROM AUTHOR]