Search

Background Hip fracture is the most common type of fracture in elderly individuals. Numerous deep learning (DL) algorithms for plain pelvic radiographs (PXRs) have been applied to improve the accuracy of hip fracture diagnosis. However, their efficacy is still undetermined. Objective The objective of this study is to develop and validate a human-algorithm integration (HAI) system to improve the accuracy of hip fracture diagnosis in a real clinical environment. Methods The HAI system with hip fracture detection ability was developed using a deep learning algorithm trained on trauma registry data and 3605 PXRs from August 2008 to December 2016. To compare their diagnostic performance before and after HAI system assistance using an independent testing dataset, 34 physicians were recruited. We analyzed the physicians’ accuracy, sensitivity, specificity, and agreement with the algorithm; we also performed subgroup analyses according to physician specialty and experience. Furthermore, we applied the HAI system in the emergency departments of different hospitals to validate its value in the real world. Results With the support of the algorithm, which achieved 91% accuracy, the diagnostic performance of physicians was significantly improved in the independent testing dataset, as was revealed by the sensitivity (physician alone, median 95%; HAI, median 99%; P Conclusions HAI currently impacts health care, and integrating this technology into emergency departments is feasible. The developed HAI system can enhance physicians’ hip fracture diagnostic performance.

Background: Sepsis is a heterogenous syndrome and individualized management strategy is the key to successful treatment. Genome wide expression profiling has been utilized for identifying subclasses of sepsis, but the clinical utility of these subclasses was limited because of the classification instability, and the lack of a robust class prediction model with extensive external validation. The study aimed to develop a parsimonious class model for the prediction of class membership and validate the model for its prognostic and predictive capability in external datasets. Methods: The Gene Expression Omnibus (GEO) and ArrayExpress databases were searched from inception to April 2020. Datasets containing whole blood gene expression profiling in adult sepsis patients were included. Autoencoder was used to extract representative features for k-means clustering. Genetic algorithms (GA) were employed to derive a parsimonious 5-gene class prediction model. The class model was then applied to external datasets (n = 780) to evaluate its prognostic and predictive performance. Findings: A total of 12 datasets involving 1613 patients were included. Two classes were identified in the discovery cohort (n = 685). Class 1 was characterized by immunosuppression with higher mortality than class 2 (21.8% [70/321] vs. 12.1% [44/364]; p < 0.01 for Chi-square test). A 5-gene class model (C14orf159, AKNA, PILRA, STOM and USP4) was developed with GA. In external validation cohorts, the 5-gene class model (AUC: 0.707; 95% CI: 0.664 – 0.750) performed better in predicting mortality than sepsis response signature (SRS) endotypes (AUC: 0.610; 95% CI: 0.521 – 0.700), and performed equivalently to the APACHE II score (AUC: 0.681; 95% CI: 0.595 – 0.767). In the dataset E-MTAB-7581, the use of hydrocortisone was associated with increased risk of mortality (OR: 3.15 [1.13, 8.82]; p = 0.029) in class 2. The effect was not statistically significant in class 1 (OR: 1.88 [0.70, 5.09]; p = 0.211). Interpretation: Our study identified two classes of sepsis that showed different mortality rates and responses to hydrocortisone therapy. Class 1 was characterized by immunosuppression with higher mortality rate than class 2. We further developed a 5-gene class model to predict class membership. Funding: The study was funded by the National Natural Science Foundation of China (Grant No. 81,901,929).

Significance: Melanoma is a deadly cancer that physicians struggle to diagnose early because they lack the knowledge to differentiate benign from malignant lesions. Deep machine learning approaches to image analysis offer promise but lack the transparency to be widely adopted as stand-alone diagnostics. Aim: We aimed to create a transparent machine learning technology (i.e., not deep learning) to discriminate melanomas from nevi in dermoscopy images and an interface for sensory cue integration. Approach: Imaging biomarker cues (IBCs) fed ensemble machine learning classifier (Eclass) training while raw images fed deep learning classifier training. We compared the areas under the diagnostic receiver operator curves. Results: Our interpretable machine learning algorithm outperformed the leading deep-learning approach 75% of the time. The user interface displayed only the diagnostic imaging biomarkers as IBCs. Conclusions: From a translational perspective, Eclass is better than convolutional machine learning diagnosis in that physicians can embrace it faster than black box outputs. Imaging biomarkers cues may be used during sensory cue integration in clinical screening. Our method may be applied to other image-based diagnostic analyses, including pathology and radiology.

The soybean cyst nematode (SCN), Heterodera glycines, is the most damaging pathogen of soybeans in the United States. To assess the severity of nematode infestations in the field, SCN egg population densities are determined. Cysts (dead females) of the nematode must be extracted from soil samples and then ground to extract the eggs within. Sucrose centrifugation commonly is used to separate debris from suspensions of extracted nematode eggs. We present a method using OptiPrep as a density gradient medium with improved separation and recovery of extracted eggs compared to the sucrose centrifugation technique. Also, computerized methods were developed to automate the identification and counting of nematode eggs from the processed samples. In one approach, a high-resolution scanner was used to take static images of extracted eggs and debris on filter papers, and a deep learning network was trained to identify and count the eggs among the debris. In the second approach, a lensless imaging setup was developed using off-the-shelf components, and the processed egg samples were passed through a microfluidic flow chip made from double-sided adhesive tape. Holographic videos were recorded of the passing eggs and debris, and the videos were reconstructed and processed by custom software program to obtain egg counts. The performance of the software programs for egg counting was characterized with SCN-infested soil collected from two farms, and the results using these methods were compared with those obtained through manual counting.

Background Fractures as a result of osteoporosis and low bone mass are common and give rise to significant clinical, personal, and economic burden. Even after a fracture occurs, high fracture risk remains widely underdiagnosed and undertreated. Common fracture risk assessment tools utilize a subset of clinical risk factors for prediction, and often require manual data entry. Furthermore, these tools predict risk over the long term and do not explicitly provide short-term risk estimates necessary to identify patients likely to experience a fracture in the next 1-2 years. Objective The goal of this study was to develop and evaluate an algorithm for the identification of patients at risk of fracture in a subsequent 1- to 2-year period. In order to address the aforementioned limitations of current prediction tools, this approach focused on a short-term timeframe, automated data entry, and the use of longitudinal data to inform the predictions. Methods Using retrospective electronic health record data from over 1,000,000 patients, we developed Crystal Bone, an algorithm that applies machine learning techniques from natural language processing to the temporal nature of patient histories to generate short-term fracture risk predictions. Similar to how language models predict the next word in a given sentence or the topic of a document, Crystal Bone predicts whether a patient’s future trajectory might contain a fracture event, or whether the signature of the patient’s journey is similar to that of a typical future fracture patient. A holdout set with 192,590 patients was used to validate accuracy. Experimental baseline models and human-level performance were used for comparison. Results The model accurately predicted 1- to 2-year fracture risk for patients aged over 50 years (area under the receiver operating characteristics curve [AUROC] 0.81). These algorithms outperformed the experimental baselines (AUROC 0.67) and showed meaningful improvements when compared to retrospective approximation of human-level performance by correctly identifying 9649 of 13,765 (70%) at-risk patients who did not receive any preventative bone-health-related medical interventions from their physicians. Conclusions These findings indicate that it is possible to use a patient’s unique medical history as it changes over time to predict the risk of short-term fracture. Validating and applying such a tool within the health care system could enable automated and widespread prediction of this risk and may help with identification of patients at very high risk of fracture.

Background Data collected by an actigraphy device worn on the wrist or waist can provide objective measurements for studies related to physical activity; however, some data may contain intervals where values are missing. In previous studies, statistical methods have been applied to impute missing values on the basis of statistical assumptions. Deep learning algorithms, however, can learn features from the data without any such assumptions and may outperform previous approaches in imputation tasks. Objective The aim of this study was to impute missing values in data using a deep learning approach. Methods To develop an imputation model for missing values in accelerometer-based actigraphy data, a denoising convolutional autoencoder was adopted. We trained and tested our deep learning–based imputation model with the National Health and Nutrition Examination Survey data set and validated it with the external Korea National Health and Nutrition Examination Survey and the Korean Chronic Cerebrovascular Disease Oriented Biobank data sets which consist of daily records measuring activity counts. The partial root mean square error and partial mean absolute error of the imputed intervals (partial RMSE and partial MAE, respectively) were calculated using our deep learning–based imputation model (zero-inflated denoising convolutional autoencoder) as well as using other approaches (mean imputation, zero-inflated Poisson regression, and Bayesian regression). Results The zero-inflated denoising convolutional autoencoder exhibited a partial RMSE of 839.3 counts and partial MAE of 431.1 counts, whereas mean imputation achieved a partial RMSE of 1053.2 counts and partial MAE of 545.4 counts, the zero-inflated Poisson regression model achieved a partial RMSE of 1255.6 counts and partial MAE of 508.6 counts, and Bayesian regression achieved a partial RMSE of 924.5 counts and partial MAE of 605.8 counts. Conclusions Our deep learning–based imputation model performed better than the other methods when imputing missing values in actigraphy data.

Motivation Deep learning use for quantitative image analysis is exponentially increasing. However, training accurate, widely deployable deep learning algorithms requires a plethora of annotated (ground truth) data. Image collections must contain not only thousands of images to provide sufficient example objects (i.e. cells), but also contain an adequate degree of image heterogeneity. Results We present a new dataset, EVICAN—Expert visual cell annotation, comprising partially annotated grayscale images of 30 different cell lines from multiple microscopes, contrast mechanisms and magnifications that is readily usable as training data for computer vision applications. With 4600 images and ∼26 000 segmented cells, our collection offers an unparalleled heterogeneous training dataset for cell biology deep learning application development. Availability and implementation The dataset is freely available (https://edmond.mpdl.mpg.de/imeji/collection/l45s16atmi6Aa4sI?q=). Using a Mask R-CNN implementation, we demonstrate automated segmentation of cells and nuclei from brightfield images with a mean average precision of 61.6 % at a Jaccard Index above 0.5.

Drug target interaction (DTI) prediction is a foundational task for in silico drug discovery, which is costly and time-consuming due to the need of experimental search over large drug compound space. Recent years have witnessed promising progress for deep learning in DTI predictions. However, the following challenges are still open: (1) the sole data-driven molecular representation learning approaches ignore the sub-structural nature of DTI, thus produce results that are less accurate and difficult to explain; (2) existing methods focus on limited labeled data while ignoring the value of massive unlabelled molecular data. We propose a Molecular Interaction Transformer (MolTrans) to address these limitations via: (1) knowledge inspired sub-structural pattern mining algorithm and interaction modeling module for more accurate and interpretable DTI prediction; (2) an augmented transformer encoder to better extract and capture the semantic relations among substructures extracted from massive unlabeled biomedical data. We evaluate MolTrans on real world data and show it improved DTI prediction performance compared to state-of-the-art baselines., Bioinformatics, 2020

BackgroundCOVID-19 has spread very rapidly, and it is important to build a system that can detect it in order to help an overwhelmed health care system. Many research studies on chest diseases rely on the strengths of deep learning techniques. Although some of these studies used state-of-the-art techniques and were able to deliver promising results, these techniques are not very useful if they can detect only one type of disease without detecting the others.ObjectiveThe main objective of this study was to achieve a fast and more accurate diagnosis of COVID-19. This study proposes a diagnostic technique that classifies COVID-19 x-ray images from normal x-ray images and those specific to 14 other chest diseases.MethodsIn this paper, we propose a novel, multilevel pipeline, based on deep learning models, to detect COVID-19 along with other chest diseases based on x-ray images. This pipeline reduces the burden of a single network to classify a large number of classes. The deep learning models used in this study were pretrained on the ImageNet dataset, and transfer learning was used for fast training. The lungs and heart were segmented from the whole x-ray images and passed onto the first classifier that checks whether the x-ray is normal, COVID-19 affected, or characteristic of another chest disease. If it is neither a COVID-19 x-ray image nor a normal one, then the second classifier comes into action and classifies the image as one of the other 14 diseases.ResultsWe show how our model uses state-of-the-art deep neural networks to achieve classification accuracy for COVID-19 along with 14 other chest diseases and normal cases based on x-ray images, which is competitive with currently used state-of-the-art models. Due to the lack of data in some classes such as COVID-19, we applied 10-fold cross-validation through the ResNet50 model. Our classification technique thus achieved an average training accuracy of 96.04% and test accuracy of 92.52% for the first level of classification (ie, 3 classes). For the second level of classification (ie, 14 classes), our technique achieved a maximum training accuracy of 88.52% and test accuracy of 66.634% by using ResNet50. We also found that when all the 16 classes were classified at once, the overall accuracy for COVID-19 detection decreased, which in the case of ResNet50 was 88.92% for training data and 71.905% for test data.ConclusionsOur proposed pipeline can detect COVID-19 with a higher accuracy along with detecting 14 other chest diseases based on x-ray images. This is achieved by dividing the classification task into multiple steps rather than classifying them collectively.

Motivation Sequence analysis is arguably a foundation of modern biology. Classic approaches to sequence analysis are based on sequence alignment, which is limited when dealing with large-scale sequence data. A dozen of alignment-free approaches have been developed to provide computationally efficient alternatives to alignment-based approaches. However, existing methods define sequence similarity based on various heuristics and can only provide rough approximations to alignment distances. Results In this article, we developed a new approach, referred to as SENSE (SiamEse Neural network for Sequence Embedding), for efficient and accurate alignment-free sequence comparison. The basic idea is to use a deep neural network to learn an explicit embedding function based on a small training dataset to project sequences into an embedding space so that the mean square error between alignment distances and pairwise distances defined in the embedding space is minimized. To the best of our knowledge, this is the first attempt to use deep learning for alignment-free sequence analysis. A large-scale experiment was performed that demonstrated that our method significantly outperformed the state-of-the-art alignment-free methods in terms of both efficiency and accuracy. Availability and implementation Open-source software for the proposed method is developed and freely available at https://www.acsu.buffalo.edu/∼yijunsun/lab/SENSE.html. Supplementary information Supplementary data are available at Bioinformatics online.

The reconstruction quality of dental computed tomography (DCT) is vulnerable to metal implants because the presence of dense metallic objects causes beam hardening and streak artifacts in the reconstructed images. These metal artifacts degrade the images and decrease the clinical usefulness of DCT. Although interpolation-based metal artifact reduction (MAR) methods have been introduced, they may not be efficient in DCT because teeth as well as metallic objects have high X-ray attenuation. In this study, we investigated an effective MAR method based on a fully convolutional network (FCN) in both sinogram and image domains. The method consisted of three main steps: (1) segmentation of the metal trace, (2) FCN-based restoration in the sinogram domain, and (3) FCN-based restoration in image domain followed by metal insertion. We performed a computational simulation and an experiment to investigate the image quality and evaluated the effectiveness of the proposed method. The results of the proposed method were compared with those obtained by the normalized MAR method and the deep learning–based MAR algorithm in the sinogram domain with respect to the root-mean-square error and the structural similarity. Our results indicate that the proposed MAR method significantly reduced the presence of metal artifacts in DCT images and demonstrated better image performance than those of the other algorithms in reducing the streak artifacts without introducing any contrast anomaly.

Cell movement in the early phase of C. elegans development is regulated by a highly complex process in which a set of rules and connections are formulated at distinct scales. Previous efforts have shown that agent-based, multi-scale modeling systems can integrate physical and biological rules and provide new avenues to study developmental systems. However, the application of these systems to model cell movement is still challenging and requires a comprehensive understanding of regulation networks at the right scales. Recent developments in deep learning and reinforcement learning provide an unprecedented opportunity to explore cell movement using 3D time-lapse images. We present a deep reinforcement learning approach within an ABM system to characterize cell movement in C. elegans embryogenesis. Our modeling system captures the complexity of cell movement patterns in the embryo and overcomes the local optimization problem encountered by traditional rule-based, ABM that uses greedy algorithms. We tested our model with two real developmental processes: the anterior movement of the Cpaaa cell via intercalation and the rearrangement of the left-right asymmetry. In the first case, model results showed that Cpaaa's intercalation is an active directional cell movement caused by the continuous effects from a longer distance, as opposed to a passive movement caused by neighbor cell movements. This is because the learning-based simulation found that a passive movement model could not lead Cpaaa to the predefined destination. In the second case, a leader-follower mechanism well explained the collective cell movement pattern. These results showed that our approach to introduce deep reinforcement learning into ABM can test regulatory mechanisms by exploring cell migration paths in a reverse engineering perspective. This model opens new doors to explore large datasets generated by live imaging., We revised the manuscript to make it clearer to follow. Please notice that the Abstract shown in this page is slightly different than that in the manuscript due to the limitation of 1920 characters in arxiv.org

ObjectiveThe present study tested the combination of an established and a validated food-choice research method (the ‘fake food buffet’) with a new food-matching technology to automate the data collection and analysis.DesignThe methodology combines fake-food image recognition using deep learning and food matching and standardization based on natural language processing. The former is specific because it uses a single deep learning network to perform both the segmentation and the classification at the pixel level of the image. To assess its performance, measures based on the standard pixel accuracy and Intersection over Union were applied. Food matching firstly describes each of the recognized food items in the image and then matches the food items with their compositional data, considering both their food names and their descriptors.ResultsThe final accuracy of the deep learning model trained on fake-food images acquired by 124 study participants and providing fifty-five food classes was 92·18 %, while the food matching was performed with a classification accuracy of 93 %.ConclusionsThe present findings are a step towards automating dietary assessment and food-choice research. The methodology outperforms other approaches in pixel accuracy, and since it is the first automatic solution for recognizing the images of fake foods, the results could be used as a baseline for possible future studies. As the approach enables a semi-automatic description of recognized food items (e.g. with respect to FoodEx2), these can be linked to any food composition database that applies the same classification and description system.

Self-interacting proteins (SIPs) play a significant role in the execution of most important molecular processes in cells, such as signal transduction, gene expression regulation, immune response and enzyme activation. Although the traditional experimental methods can be used to generate SIPs data, it is very expensive and time-consuming based only on biological technique. Therefore, it is important and urgent to develop an efficient computational method for SIPs detection. In this study, we present a novel SIPs identification method based on machine learning technology by combing the Zernike Moments (ZMs) descriptor on Position Specific Scoring Matrix (PSSM) with Probabilistic Classification Vector Machines (PCVM) and Stacked Sparse Auto-Encoder (SSAE). More specifically, an efficient feature extraction technique called ZMs is firstly utilized to generate feature vectors on Position Specific Scoring Matrix (PSSM); Then, Deep neural network is employed for reducing the feature dimensions and noise; Finally, the Probabilistic Classification Vector Machine is used to execute the classification. The prediction performance of the proposed method is evaluated on S.erevisiae and Human SIPs datasets via cross-validation. The experimental results indicate that the proposed method can achieve good accuracies of 92.55% and 97.47%, respectively. To further evaluate the advantage of our scheme for SIPs prediction, we also compared the PCVM classifier with the Support Vector Machine (SVM) and other existing techniques on the same data sets. Comparison results reveal that the proposed strategy is outperforms other methods and could be a used tool for identifying SIPs.

Pathological evaluation of tumor tissue is pivotal for diagnosis in cancer patients and automated image analysis approaches have great potential to increase precision of diagnosis and help reduce human error. In this study, we utilize several computational methods based on convolutional neural networks (CNN) and build a stand-alone pipeline to effectively classify different histopathology images across different types of cancer. In particular, we demonstrate the utility of our pipeline to discriminate between two subtypes of lung cancer, four biomarkers of bladder cancer, and five biomarkers of breast cancer. In addition, we apply our pipeline to discriminate among four immunohistochemistry (IHC) staining scores of bladder and breast cancers. Our classification pipeline includes a basic CNN architecture, Google's Inceptions with three training strategies, and an ensemble of two state-of-the-art algorithms, Inception and ResNet. Training strategies include training the last layer of Google's Inceptions, training the network from scratch, and fine-tunning the parameters for our data using two pre-trained version of Google's Inception architectures, Inception-V1 and Inception-V3. We demonstrate the power of deep learning approaches for identifying cancer subtypes, and the robustness of Google's Inceptions even in presence of extensive tumor heterogeneity. On average, our pipeline achieved accuracies of 100%, 92%, 95%, and 69% for discrimination of various cancer tissues, subtypes, biomarkers, and scores, respectively. Our pipeline and related documentation is freely available at https://github.com/ih-_lab/CNN_Smoothie., Highlights • This work presents the power of deep learning approaches for classifying pathology images in presence of extensive tumor heterogeneity. • A basic CNN architecture and several state-of-the-art image classification architectures are compared in an open source pipeline. • The pipeline is applicable to wide rang of histopathology images.

Background: To reduce the high incidence and mortality of gastric cancer (GC), we aimed to develop deep learning-based models to assist in predicting the diagnosis and overall survival (OS) of GC patients using pathological images. Methods: 2333 hematoxylin and eosin-stained pathological pictures of 1037 GC patients were collected from two cohorts to develop our algorithms, Renmin Hospital of Wuhan University (RHWU) and the Cancer Genome Atlas (TCGA). Additionally, we gained 175 digital pictures of 91 GC patients from National Human Genetic Resources Sharing Service Platform (NHGRP), served as the independent external validation set. Two models were developed using artificial intelligence (AI), one named GastroMIL for diagnosing GC, and the other named MIL-GC for predicting outcome of GC. Findings: The discriminatory power of GastroMIL achieved accuracy 0.920 in the external validation set, superior to that of the junior pathologist and comparable to that of expert pathologists. In the prognostic model, C-indices for survival prediction of internal and external validation sets were 0.671 and 0.657, respectively. Moreover, the risk score output by MIL-GC in the external validation set was proved to be a strong predictor of OS both in the univariate (HR = 2.414, P

Background COVID-19 is caused by the SARS-CoV-2 virus and has strikingly heterogeneous clinical manifestations, with most individuals contracting mild disease but a substantial minority experiencing fulminant cardiopulmonary symptoms or death. The clinical covariates and the laboratory tests performed on a patient provide robust statistics to guide clinical treatment. Deep learning approaches on a data set of this nature enable patient stratification and provide methods to guide clinical treatment. Objective Here, we report on the development and prospective validation of a state-of-the-art machine learning model to provide mortality prediction shortly after confirmation of SARS-CoV-2 infection in the Mayo Clinic patient population. Methods We retrospectively constructed one of the largest reported and most geographically diverse laboratory information system and electronic health record of COVID-19 data sets in the published literature, which included 11,807 patients residing in 41 states of the United States of America and treated at medical sites across 5 states in 3 time zones. Traditional machine learning models were evaluated independently as well as in a stacked learner approach by using AutoGluon, and various recurrent neural network architectures were considered. The traditional machine learning models were implemented using the AutoGluon-Tabular framework, whereas the recurrent neural networks utilized the TensorFlow Keras framework. We trained these models to operate solely using routine laboratory measurements and clinical covariates available within 72 hours of a patient’s first positive COVID-19 nucleic acid test result. Results The GRU-D recurrent neural network achieved peak cross-validation performance with 0.938 (SE 0.004) as the area under the receiver operating characteristic (AUROC) curve. This model retained strong performance by reducing the follow-up time to 12 hours (0.916 [SE 0.005] AUROC), and the leave-one-out feature importance analysis indicated that the most independently valuable features were age, Charlson comorbidity index, minimum oxygen saturation, fibrinogen level, and serum iron level. In the prospective testing cohort, this model provided an AUROC of 0.901 and a statistically significant difference in survival (P Conclusions Our deep learning approach using GRU-D provides an alert system to flag mortality for COVID-19–positive patients by using clinical covariates and laboratory values within a 72-hour window after the first positive nucleic acid test result.

Background When using a smartwatch to obtain electrocardiogram (ECG) signals from multiple leads, the device has to be placed on different parts of the body sequentially. The ECG signals measured from different leads are asynchronous. Artificial intelligence (AI) models for asynchronous ECG signals have barely been explored. Objective We aimed to develop an AI model for detecting acute myocardial infarction using asynchronous ECGs and compare its performance with that of the automatic ECG interpretations provided by a commercial ECG analysis software. We sought to evaluate the feasibility of implementing multiple lead–based AI-enabled ECG algorithms on smartwatches. Moreover, we aimed to determine the optimal number of leads for sufficient diagnostic power. Methods We extracted ECGs recorded within 24 hours from each visit to the emergency room of Ajou University Medical Center between June 1994 and January 2018 from patients aged 20 years or older. The ECGs were labeled on the basis of whether a diagnostic code corresponding to acute myocardial infarction was entered. We derived asynchronous ECG lead sets from standard 12-lead ECG reports and simulated a situation similar to the sequential recording of ECG leads via smartwatches. We constructed an AI model based on residual networks and self-attention mechanisms by randomly masking each lead channel during the training phase and then testing the model using various targeting lead sets with the remaining lead channels masked. Results The performance of lead sets with 3 or more leads compared favorably with that of the automatic ECG interpretations provided by a commercial ECG analysis software, with 8.1%-13.9% gain in sensitivity when the specificity was matched. Our results indicate that multiple lead-based AI-enabled ECG algorithms can be implemented on smartwatches. Model performance generally increased as the number of leads increased (12-lead sets: area under the receiver operating characteristic curve [AUROC] 0.880; 4-lead sets: AUROC 0.858, SD 0.008; 3-lead sets: AUROC 0.845, SD 0.011; 2-lead sets: AUROC 0.813, SD 0.018; single-lead sets: AUROC 0.768, SD 0.001). Considering the short amount of time needed to measure additional leads, measuring at least 3 leads—ideally more than 4 leads—is necessary for minimizing the risk of failing to detect acute myocardial infarction occurring in a certain spatial location or direction. Conclusions By developing an AI model for detecting acute myocardial infarction with asynchronous ECG lead sets, we demonstrated the feasibility of multiple lead-based AI-enabled ECG algorithms on smartwatches for automated diagnosis of cardiac disorders. We also demonstrated the necessity of measuring at least 3 leads for accurate detection. Our results can be used as reference for the development of other AI models using sequentially measured asynchronous ECG leads via smartwatches for detecting various cardiac disorders.

Background Dental diseases can be prevented through the management of dental plaques. Dental plaque can be identified using the light-induced fluorescence (LIF) technique that emits light at 405 nm. The LIF technique is more convenient than the commercial technique using a disclosing agent, but the result may vary for each individual as it still requires visual identification. Objective The objective of this study is to introduce and validate a deep learning–based oral hygiene monitoring system that makes it easy to identify dental plaques at home. Methods We developed a LIF-based system consisting of a device that can visually identify dental plaques and a mobile app that displays the location and area of dental plaques on oral images. The mobile app is programmed to automatically determine the location and distribution of dental plaques using a deep learning–based algorithm and present the results to the user as time series data. The mobile app is also built with convergence of naive and web applications so that the algorithm is executed on a cloud server to efficiently distribute computing resources. Results The location and distribution of users’ dental plaques could be identified via the hand-held LIF device or mobile app. The color correction filter in the device was developed using a color mixing technique. The mobile app was built as a hybrid app combining the functionalities of a native application and a web application. Through the scrollable WebView on the mobile app, changes in the time series of dental plaque could be confirmed. The algorithm for dental plaque detection was implemented to run on Amazon Web Services for object detection by single shot multibox detector and instance segmentation by Mask region-based convolutional neural network. Conclusions This paper shows that the system can be used as a home oral care product for timely identification and management of dental plaques. In the future, it is expected that these products will significantly reduce the social costs associated with dental diseases.

Motivation Transcription factor (TF) binds to the promoter region of a gene to control gene expression. Identifying precise TF binding sites (TFBSs) is essential for understanding the detailed mechanisms of TF-mediated gene regulation. However, there is a shortage of computational approach that can deliver single base pair resolution prediction of TFBS. Results In this paper, we propose DeepSNR, a Deep Learning algorithm for predicting TF binding location at Single Nucleotide Resolution de novo from DNA sequence. DeepSNR adopts a novel deconvolutional network (deconvNet) model and is inspired by the similarity to image segmentation by deconvNet. The proposed deconvNet architecture is constructed on top of ‘DeepBind’ and we trained the entire model using TF-specific data from ChIP-exonuclease (ChIP-exo) experiments. DeepSNR has been shown to outperform motif search–based methods for several evaluation metrics. We have also demonstrated the usefulness of DeepSNR in the regulatory analysis of TFBS as well as in improving the TFBS prediction specificity using ChIP-seq data. Availability and implementation DeepSNR is available open source in the GitHub repository (https://github.com/sirajulsalekin/DeepSNR) Supplementary information Supplementary data are available at Bioinformatics online.

Background Over half a million individuals are diagnosed with head and neck cancer each year globally. Radiotherapy is an important curative treatment for this disease, but it requires manual time to delineate radiosensitive organs at risk. This planning process can delay treatment while also introducing interoperator variability, resulting in downstream radiation dose differences. Although auto-segmentation algorithms offer a potentially time-saving solution, the challenges in defining, quantifying, and achieving expert performance remain. Objective Adopting a deep learning approach, we aim to demonstrate a 3D U-Net architecture that achieves expert-level performance in delineating 21 distinct head and neck organs at risk commonly segmented in clinical practice. Methods The model was trained on a data set of 663 deidentified computed tomography scans acquired in routine clinical practice and with both segmentations taken from clinical practice and segmentations created by experienced radiographers as part of this research, all in accordance with consensus organ at risk definitions. Results We demonstrated the model’s clinical applicability by assessing its performance on a test set of 21 computed tomography scans from clinical practice, each with 21 organs at risk segmented by 2 independent experts. We also introduced surface Dice similarity coefficient, a new metric for the comparison of organ delineation, to quantify the deviation between organ at risk surface contours rather than volumes, better reflecting the clinical task of correcting errors in automated organ segmentations. The model’s generalizability was then demonstrated on 2 distinct open-source data sets, reflecting different centers and countries to model training. Conclusions Deep learning is an effective and clinically applicable technique for the segmentation of the head and neck anatomy for radiotherapy. With appropriate validation studies and regulatory approvals, this system could improve the efficiency, consistency, and safety of radiotherapy pathways.

Objectives: To evaluate image quality and lesion detection capabilities of low-dose (LD) portal venous phase whole-body computed tomography (CT) using deep learning image reconstruction (DLIR). Methods: The study cohort of 59 consecutive patients (mean age, 67.2 years) who underwent whole-body LD CT and a prior standard-dose (SD) CT reconstructed with hybrid iterative reconstruction (SD-IR) within one year for surveillance of malignancy were assessed. The LD CT images were reconstructed with hybrid iterative reconstruction of 40% (LD-IR) and DLIR (LD-DLIR). The radiologists independently evaluated image quality (5-point scale) and lesion detection. Attenuation values in Hounsfield units (HU) of the liver, pancreas, spleen, abdominal aorta, and portal vein; the background noise and signal-to-noise ratio (SNR) of the liver, pancreas, and spleen were calculated. Qualitative and quantitative parameters were compared between the SD-IR, LD-IR, and LD-DLIR images. The CT dose-index volumes (CTDIvol) and dose-length product (DLP) were compared between SD and LD scans. Results: The image quality and lesion detection rate of the LD-DLIR was comparable to the SD-IR. The image quality was significantly better in SD-IR than in LD-IR (p < 0.017). The attenuation values of all anatomical structures were comparable between the SD-IR and LD-DLIR (p = 0.28–0.96). However, background noise was significantly lower in the LD-DLIR (p < 0.001) and resulted in improved SNRs (p < 0.001) compared to the SD-IR and LD-IR images. The mean CTDIvol and DLP were significantly lower in the LD (2.9 mGy and 216.2 mGy•cm) than in the SD (13.5 mGy and 1011.6 mGy•cm) (p < 0.0001). Conclusion: LD CT images reconstructed with DLIR enable radiation dose reduction of >75% while maintaining image quality and lesion detection rate and superior SNR in comparison to SD-IR. Advances in knowledge: Deep learning image reconstruction algorithm enables around 80% reduction in radiation dose while maintaining the image quality and lesion detection compared to standard-dose whole-body CT.

Motivation Progress in 3D electron microscopy (EM) imaging has greatly facilitated neuroscience research in high-throughput data acquisition. Correspondingly, high-throughput automated image analysis methods are necessary to work on par with the speed of data being produced. One such example is the need for automated EM image segmentation for neurite reconstruction. However, the efficiency and reliability of current methods are still lagging far behind human performance. Results Here, we propose DeepEM3D, a deep learning method for segmenting 3D anisotropic brain electron microscopy images. In this method, the deep learning model can efficiently build feature representation and incorporate sufficient multi-scale contextual information. We propose employing a combination of novel boundary map generation methods with optimized model ensembles to address the inherent challenges of segmenting anisotropic images. We evaluated our method by participating in the 3D segmentation of neurites in EM images (SNEMI3D) challenge. Our submission is ranked #1 on the current leaderboard as of Oct 15, 2016. More importantly, our result was very close to human-level performance in terms of the challenge evaluation metric: namely, a Rand error of 0.06015 versus the human value of 0.05998. Availability and Implementation The code is available at https://github.com/divelab/deepem3d/ Supplementary information Supplementary data are available at Bioinformatics online.

Background: The spatial distributions of different types of cells could reveal a cancer cell's growth pattern, its relationships with the tumor microenvironment and the immune response of the body, all of which represent key “hallmarks of cancer”. However, the process by which pathologists manually recognize and localize all the cells in pathology slides is extremely labor intensive and error prone. Methods: In this study, we developed an automated cell type classification pipeline, ConvPath, which includes nuclei segmentation, convolutional neural network-based tumor cell, stromal cell, and lymphocyte classification, and extraction of tumor microenvironment-related features for lung cancer pathology images. To facilitate users in leveraging this pipeline for their research, all source scripts for ConvPath software are available at https://qbrc.swmed.edu/projects/cnn/. Findings: The overall classification accuracy was 92.9% and 90.1% in training and independent testing datasets, respectively. By identifying cells and classifying cell types, this pipeline can convert a pathology image into a “spatial map” of tumor, stromal and lymphocyte cells. From this spatial map, we can extract features that characterize the tumor micro-environment. Based on these features, we developed an image feature-based prognostic model and validated the model in two independent cohorts. The predicted risk group serves as an independent prognostic factor, after adjusting for clinical variables that include age, gender, smoking status, and stage. Interpretation: The analysis pipeline developed in this study could convert the pathology image into a “spatial map” of tumor cells, stromal cells and lymphocytes. This could greatly facilitate and empower comprehensive analysis of the spatial organization of cells, as well as their roles in tumor progression and metastasis. Keywords: Deep learning, Convolutional neural network, Lung adenocarcinoma, Pathology image, Cell distribution and interaction, Prognosis

In the context of epilepsy monitoring, electroencephalography (EEG) remains the modality of choice. Functional near-infrared spectroscopy (fNIRS) is a relatively innovative modality that cannot only characterize hemodynamic profiles of seizures but also allow for long-term recordings. We employ deep learning methods to investigate the benefits of integrating fNIRS measures for seizure detection. We designed a deep recurrent neural network with long short-term memory units and subsequently validated it using the CHBMIT scalp EEG database—a compendium of 896 h of surface EEG seizure recordings. After validating our network using EEG, fNIRS, and multimodal data comprising a corpus of 89 seizures from 40 refractory epileptic patients was used as model input to evaluate the integration of fNIRS measures. Following heuristic hyperparameter optimization, multimodal EEG-fNIRS data provide superior performance metrics (sensitivity and specificity of 89.7% and 95.5%, respectively) in a seizure detection task, with low generalization errors and loss. False detection rates are generally low, with 11.8% and 5.6% for EEG and multimodal data, respectively. Employing multimodal neuroimaging, particularly EEG-fNIRS, in epileptic patients, can enhance seizure detection performance. Furthermore, the neural network model proposed and characterized herein offers a promising framework for future multimodal investigations in seizure detection and prediction.

Motivation: Many researchers with domain expertise are unable to easily apply machine learning to their bioinformatics data due to a lack of machine learning and/or coding expertise. Methods that have been proposed thus far to automate machine learning mostly require programming experience as well as expert knowledge to tune and apply the algorithms correctly. Here, we study a method of automating biomedical data science using a web-based platform that uses AI to recommend model choices and conduct experiments. We have two goals in mind: first, to make it easy to construct sophisticated models of biomedical processes; and second, to provide a fully automated AI agent that can choose and conduct promising experiments for the user, based on the user's experiments as well as prior knowledge. To validate this framework, we experiment with hundreds of classification problems, comparing to state-of-the-art, automated approaches. Finally, we use this tool to develop predictive models of septic shock in critical care patients. Results: We find that matrix factorization-based recommendation systems outperform meta-learning methods for automating machine learning. This result mirrors the results of earlier recommender systems research in other domains. The proposed AI is competitive with state-of-the-art automated machine learning methods in terms of choosing optimal algorithm configurations for datasets. In our application to prediction of septic shock, the AI-driven analysis produces a competent machine learning model (AUROC 0.85 +/- 0.02) that performs on par with state-of-the-art deep learning results for this task, with much less computational effort., Comment: 17 pages, 8 figures. this version fixes link to pennai in abstract

Background Multimodal wearable technologies have brought forward wide possibilities in human activity recognition, and more specifically personalized monitoring of eating habits. The emerging challenge now is the selection of most discriminative information from high-dimensional data collected from multiple sources. The available fusion algorithms with their complex structure are poorly adopted to the computationally constrained environment which requires integrating information directly at the source. As a result, more simple low-level fusion methods are needed. Objective In the absence of a data combining process, the cost of directly applying high-dimensional raw data to a deep classifier would be computationally expensive with regard to the response time, energy consumption, and memory requirement. Taking this into account, we aimed to develop a data fusion technique in a computationally efficient way to achieve a more comprehensive insight of human activity dynamics in a lower dimension. The major objective was considering statistical dependency of multisensory data and exploring intermodality correlation patterns for different activities. Methods In this technique, the information in time (regardless of the number of sources) is transformed into a 2D space that facilitates classification of eating episodes from others. This is based on a hypothesis that data captured by various sensors are statistically associated with each other and the covariance matrix of all these signals has a unique distribution correlated with each activity which can be encoded on a contour representation. These representations are then used as input of a deep model to learn specific patterns associated with specific activity. Results In order to show the generalizability of the proposed fusion algorithm, 2 different scenarios were taken into account. These scenarios were different in terms of temporal segment size, type of activity, wearable device, subjects, and deep learning architecture. The first scenario used a data set in which a single participant performed a limited number of activities while wearing the Empatica E4 wristband. In the second scenario, a data set related to the activities of daily living was used where 10 different participants wore inertial measurement units while performing a more complex set of activities. The precision metric obtained from leave-one-subject-out cross-validation for the second scenario reached 0.803. The impact of missing data on performance degradation was also evaluated. Conclusions To conclude, the proposed fusion technique provides the possibility of embedding joint variability information over different modalities in just a single 2D representation which results in obtaining a more global view of different aspects of daily human activities at hand, and yet preserving the desired performance level in activity recognition.

BackgroundDespite excellent prediction performance, noninterpretability has undermined the value of applying deep-learning algorithms in clinical practice. To overcome this limitation, attention mechanism has been introduced to clinical research as an explanatory modeling method. However, potential limitations of using this attractive method have not been clarified to clinical researchers. Furthermore, there has been a lack of introductory information explaining attention mechanisms to clinical researchers.ObjectiveThe aim of this study was to introduce the basic concepts and design approaches of attention mechanisms. In addition, we aimed to empirically assess the potential limitations of current attention mechanisms in terms of prediction and interpretability performance.MethodsFirst, the basic concepts and several key considerations regarding attention mechanisms were identified. Second, four approaches to attention mechanisms were suggested according to a two-dimensional framework based on the degrees of freedom and uncertainty awareness. Third, the prediction performance, probability reliability, concentration of variable importance, consistency of attention results, and generalizability of attention results to conventional statistics were assessed in the diabetic classification modeling setting. Fourth, the potential limitations of attention mechanisms were considered.ResultsPrediction performance was very high for all models. Probability reliability was high in models with uncertainty awareness. Variable importance was concentrated in several variables when uncertainty awareness was not considered. The consistency of attention results was high when uncertainty awareness was considered. The generalizability of attention results to conventional statistics was poor regardless of the modeling approach.ConclusionsThe attention mechanism is an attractive technique with potential to be very promising in the future. However, it may not yet be desirable to rely on this method to assess variable importance in clinical settings. Therefore, along with theoretical studies enhancing attention mechanisms, more empirical studies investigating potential limitations should be encouraged.

Background Identifying which individuals will develop tuberculosis (TB) remains an unresolved problem due to few animal models and computational approaches that effectively address its heterogeneity. To meet these shortcomings, we show that Diversity Outbred (DO) mice reflect human-like genetic diversity and develop human-like lung granulomas when infected with Mycobacterium tuberculosis (M.tb) . Methods Following M.tb infection, a “supersusceptible” phenotype develops in approximately one-third of DO mice characterized by rapid morbidity and mortality within 8 weeks. These supersusceptible DO mice develop lung granulomas patterns akin to humans. This led us to utilize deep learning to identify supersusceptibility from hematoxylin & eosin (H&E) lung tissue sections utilizing only clinical outcomes (supersusceptible or not-supersusceptible) as labels. Findings The proposed machine learning model diagnosed supersusceptibility with high accuracy (91.50 ± 4.68%) compared to two expert pathologists using H&E stained lung sections (94.95% and 94.58%). Two non-experts used the imaging biomarker to diagnose supersusceptibility with high accuracy (88.25% and 87.95%) and agreement (96.00%). A board-certified veterinary pathologist (GB) examined the imaging biomarker and determined the model was making diagnostic decisions using a form of granuloma necrosis (karyorrhectic and pyknotic nuclear debris). This was corroborated by one other board-certified veterinary pathologist. Finally, the imaging biomarker was quantified, providing a novel means to convert visual patterns within granulomas to data suitable for statistical analyses. Implications Overall, our results have translatable implication to improve our understanding of TB and also to the broader field of computational pathology in which clinical outcomes alone can drive automatic identification of interpretable imaging biomarkers, knowledge discovery, and validation of existing clinical biomarkers. Funding National Institutes of Health and American Lung Association.

Background Diagnosis of rib fractures plays an important role in identifying trauma severity. However, quickly and precisely identifying the rib fractures in a large number of CT images with increasing number of patients is a tough task, which is also subject to the qualification of radiologist. We aim at a clinically applicable automatic system for rib fracture detection and segmentation from CT scans. Methods A total of 7,473 annotated traumatic rib fractures from 900 patients in a single center were enrolled into our dataset, named RibFrac Dataset, which were annotated with a human-in-the-loop labeling procedure. We developed a deep learning model, named FracNet, to detect and segment rib fractures. 720, 60 and 120 patients were randomly split as training cohort, tuning cohort and test cohort, respectively. Free-Response ROC (FROC) analysis was used to evaluate the sensitivity and false positives of the detection performance, and Intersection-over-Union (IoU) and Dice Coefficient (Dice) were used to evaluate the segmentation performance of predicted rib fractures. Observer studies, including independent human-only study and human-collaboration study, were used to benchmark the FracNet with human performance and evaluate its clinical applicability. A annotated subset of RibFrac Dataset, including 420 for training, 60 for tuning and 120 for test, as well as our code for model training and evaluation, was open to research community to facilitate both clinical and engineering research. Findings Our method achieved a detection sensitivity of 92.9% with 5.27 false positives per scan and a segmentation Dice of 71.5%on the test cohort. Human experts achieved much lower false positives per scan, while underperforming the deep neural networks in terms of detection sensitivities with longer time in diagnosis. With human-computer collobration, human experts achieved higher detection sensitivities than human-only or computer-only diagnosis. Interpretation The proposed FracNet provided increasing detection sensitivity of rib fractures with significantly decreased clinical time consumed, which established a clinically applicable method to assist the radiologist in clinical practice. Funding A full list of funding bodies that contributed to this study can be found in the Acknowledgements section. The funding sources played no role in the study design; collection, analysis, and interpretation of data; writing of the report; or decision to submit the article for publication .

Motivation: Recently, confocal light sheet microscopy has enabled high-throughput acquisition of whole mouse brain 3D images at the micron scale resolution. This poses the unprecedented challenge of creating accurate digital maps of the whole set of cells in a brain. Results: We introduce a fast and scalable algorithm for fully automated cell identification. We obtained the whole digital map of Purkinje cells in mouse cerebellum consisting of a set of 3D cell center coordinates. The method is accurate and we estimated an F1 measure of 0.96 using 56 representative volumes, totaling 1.09 GVoxel and containing 4138 manually annotated soma centers. Availability and implementation: Source code and its documentation are available at http://bcfind.dinfo.unifi.it/. The whole pipeline of methods is implemented in Python and makes use of Pylearn2 and modified parts of Scikit-learn. Brain images are available on request. Contact: paolo.frasconi@unifi.it Supplementary information: Supplementary data are available at Bioinformatics online.

Motivation: Alternative splicing (AS) is a regulated process that directs the generation of different transcripts from single genes. A computational model that can accurately predict splicing patterns based on genomic features and cellular context is highly desirable, both in understanding this widespread phenomenon, and in exploring the effects of genetic variations on AS. Methods: Using a deep neural network, we developed a model inferred from mouse RNA-Seq data that can predict splicing patterns in individual tissues and differences in splicing patterns across tissues. Our architecture uses hidden variables that jointly represent features in genomic sequences and tissue types when making predictions. A graphics processing unit was used to greatly reduce the training time of our models with millions of parameters. Results: We show that the deep architecture surpasses the performance of the previous Bayesian method for predicting AS patterns. With the proper optimization procedure and selection of hyperparameters, we demonstrate that deep architectures can be beneficial, even with a moderately sparse dataset. An analysis of what the model has learned in terms of the genomic features is presented. Contact: frey@psi.toronto.edu Supplementary information: Supplementary data are available at Bioinformatics online.

Aging biomarkers are the qualitative and quantitative indicators of the aging processes of the human body. Estimation of biological age is important for assessing the physiological state of an organism. The advent of machine learning lead to the development of the many age predictors commonly referred to as the "aging clocks" varying in biological relevance, ease of use, cost, actionability, interpretability, and applications. Here we present and investigate a novel non-invasive class of visual photographic biomarkers of aging. We developed a simple and accurate predictor of chronological age using just the anonymized images of eye corners called the PhotoAgeClock. Deep neural networks were trained on 8414 anonymized high-resolution images of eye corners labeled with the correct chronological age. For people within the age range of 20 to 80 in a specific population, the model was able to achieve a mean absolute error of 2.3 years and 95% Pearson and Spearman correlation.

The objective of this work is to develop a computer-aided diagnostic system for early diagnosis of prostate cancer. The presented system integrates both clinical biomarkers (prostate-specific antigen) and extracted features from diffusion-weighted magnetic resonance imaging collected at multiple b values. The presented system performs 3 major processing steps. First, prostate delineation using a hybrid approach that combines a level-set model with nonnegative matrix factorization. Second, estimation and normalization of diffusion parameters, which are the apparent diffusion coefficients of the delineated prostate volumes at different b values followed by refinement of those apparent diffusion coefficients using a generalized Gaussian Markov random field model. Then, construction of the cumulative distribution functions of the processed apparent diffusion coefficients at multiple b values. In parallel, a K-nearest neighbor classifier is employed to transform the prostate-specific antigen results into diagnostic probabilities. Finally, those prostate-specific antigen–based probabilities are integrated with the initial diagnostic probabilities obtained using stacked nonnegativity constraint sparse autoencoders that employ apparent diffusion coefficient–cumulative distribution functions for better diagnostic accuracy. Experiments conducted on 18 diffusion-weighted magnetic resonance imaging data sets achieved 94.4% diagnosis accuracy (sensitivity = 88.9% and specificity = 100%), which indicate the promising results of the presented computer-aided diagnostic system.

OBJECTIVE: To train a generic deep learning software (DLS) to classify breast cancer on ultrasound images and to compare its performance to human readers with variable breast imaging experience. METHODS: In this retrospective study, all breast ultrasound examinations from January 1, 2014 to December 31, 2014 at our institution were reviewed. Patients with post-surgical scars, initially indeterminate, or malignant lesions with histological diagnoses or 2-year follow-up were included. The DLS was trained with 70% of the images, and the remaining 30% were used to validate the performance. Three readers with variable expertise also evaluated the validation set (radiologist, resident, medical student). Diagnostic accuracy was assessed with a receiver operating characteristic analysis. RESULTS: 82 patients with malignant and 550 with benign lesions were included. Time needed for training was 7 min (DLS). Evaluation time for the test data set were 3.7 s (DLS) and 28, 22 and 25 min for human readers (decreasing experience). Receiver operating characteristic analysis revealed non-significant differences (p-values 0.45–0.47) in the area under the curve of 0.84 (DLS), 0.88 (experienced and intermediate readers) and 0.79 (inexperienced reader). CONCLUSION: DLS may aid diagnosing cancer on breast ultrasound images with an accuracy comparable to radiologists, and learns better and faster than a human reader with no prior experience. Further clinical trials with dedicated algorithms are warranted. ADVANCES IN KNOWLEDGE: DLS can be trained classify cancer on breast ultrasound images high accuracy even with comparably few training cases. The fast evaluation speed makes real-time image analysis feasible.

Motivation An accurate characterization of transcription factor (TF)-DNA affinity landscape is crucial to a quantitative understanding of the molecular mechanisms underpinning endogenous gene regulation. While recent advances in biotechnology have brought the opportunity for building binding affinity prediction methods, the accurate characterization of TF-DNA binding affinity landscape still remains a challenging problem. Results Here we propose a novel sequence embedding approach for modeling the transcription factor binding affinity landscape. Our method represents DNA binding sequences as a hidden Markov model which captures both position specific information and long-range dependency in the sequence. A cornerstone of our method is a novel message passing-like embedding algorithm, called Sequence2Vec, which maps these hidden Markov models into a common nonlinear feature space and uses these embedded features to build a predictive model. Our method is a novel combination of the strength of probabilistic graphical models, feature space embedding and deep learning. We conducted comprehensive experiments on over 90 large-scale TF-DNA datasets which were measured by different high-throughput experimental technologies. Sequence2Vec outperforms alternative machine learning methods as well as the state-of-the-art binding affinity prediction methods. Availability and implementation Our program is freely available at https://github.com/ramzan1990/sequence2vec. Supplementary information Supplementary data are available at Bioinformatics online.

Background: An increasing number of doctor reviews are being generated by patients on the internet. These reviews address a diverse set of topics (features), including wait time, office staff, doctor’s skills, and bedside manners. Most previous work on automatic analysis of Web-based customer reviews assumes that (1) product features are described unambiguously by a small number of keywords, for example, battery for phones and (2) the opinion for each feature has a positive or negative sentiment. However, in the domain of doctor reviews, this setting is too restrictive: a feature such as visit duration for doctor reviews may be expressed in many ways and does not necessarily have a positive or negative sentiment. Objective: This study aimed to adapt existing and propose novel text classification methods on the domain of doctor reviews. These methods are evaluated on their accuracy to classify a diverse set of doctor review features. Methods: We first manually examined a large number of reviews to extract a set of features that are frequently mentioned in the reviews. Then we proposed a new algorithm that goes beyond bag-of-words or deep learning classification techniques by leveraging natural language processing (NLP) tools. Specifically, our algorithm automatically extracts dependency tree patterns and uses them to classify review sentences. Results: We evaluated several state-of-the-art text classification algorithms as well as our dependency tree–based classifier algorithm on a real-world doctor review dataset. We showed that methods using deep learning or NLP techniques tend to outperform traditional bag-of-words methods. In our experiments, the 2 best methods used NLP techniques; on average, our proposed classifier performed 2.19% better than an existing NLP-based method, but many of its predictions of specific opinions were incorrect. Conclusions: We conclude that it is feasible to classify doctor reviews. Automatically classifying these reviews would allow patients to easily search for doctors based on their personal preference criteria.