A transformer model for learning spatiotemporal contextual representation in fMRI data.

Representation learning is a core component in data-driven modeling of various complex phenomena. Learning a contextually informative representation can especially benefit the analysis of fMRI data because of the complexities and dynamic dependencies present in such datasets. In this work, we propose a framework based on transformer models to learn an embedding of the fMRI data by taking the spatiotemporal contextual information in the data into account. This approach takes the multivariate BOLD time series of the regions of the brain as well as their functional connectivity network simultaneously as the input to create a set of meaningful features that can in turn be used in various downstream tasks such as classification, feature extraction, and statistical analysis. The proposed spatiotemporal framework uses the attention mechanism as well as the graph convolution neural network to jointly inject the contextual information regarding the dynamics in time series data and their connectivity into the representation. We demonstrate the benefits of this framework by applying it to two resting-state fMRI datasets, and provide further discussion on various aspects and advantages of it over a number of other commonly adopted architectures.

Dialysis Patients' Social Networks and Living Donation Offers.

Rationale & Objective: Most living kidney donors are members of a hemodialysis patient's social network. Network members are divided into core members, those strongly connected to the patient and other members; and peripheral members, those weakly connected to the patient and other members. We identify how many hemodialysis patients' network members offered to become kidney donors, whether these offers were from core or peripheral network members, and whose offers the patients accepted. Study Design: A cross-sectional interviewer-administered hemodialysis patient social network survey. Setting & Participants: Prevalent hemodialysis patients in 2 facilities. Predictors: Network size and constraint, a donation from a peripheral network member. Outcomes: Number of living donor offers, accepting an offer. Analytical Approach: We performed egocentric network analyses for all participants. Poisson regression models evaluated associations between network measures and number of offers. Logistic regression models determined the associations between network factors and accepting a donation offer. Results: The mean age of the 106 participants was 60 years. Forty-five percent were female, and 75% self-identified as Black. Fifty-two percent of participants received at least one living donor offer (range 1-6); 42% of the offers were from peripheral members. Participants with larger networks received more offers (incident rate ratio [IRR], 1.26; 95% CI, 1.12-1.42; P = 0.001), including networks with more peripheral members (constraint, IRR, 0.97; 95% CI, 0.96-0.98; P < 0.001). Participants who received a peripheral member offer had 3.6 times greater odds of accepting an offer (OR, 3.56; 95% CI, 1.15-10.8; P = 0.02) than those who did not receive a peripheral member offer. Limitations: A small sample of only hemodialysis patients. Conclusions: Most participants received at least one living donor offer, often from peripheral network members. Future living donor interventions should focus on both core and peripheral network members.

Discovering research articles containing evolutionary timetrees by machine learning.

MOTIVATION: Timetrees depict evolutionary relationships between species and the geological times of their divergence. Hundreds of research articles containing timetrees are published in scientific journals every year. The TimeTree (TT) project has been manually locating, curating and synthesizing timetrees from these articles for almost two decades into a TimeTree of Life, delivered through a unique, user-friendly web interface (timetree.org). The manual process of finding articles containing timetrees is becoming increasingly expensive and time-consuming. So, we have explored the effectiveness of text-mining approaches and developed optimizations to find research articles containing timetrees automatically. RESULTS: We have developed an optimized machine learning system to determine if a research article contains an evolutionary timetree appropriate for inclusion in the TT resource. We found that BERT classification fine-tuned on whole-text articles achieved an F1 score of 0.67, which we increased to 0.88 by text-mining article excerpts surrounding the mentioning of figures. The new method is implemented in the TimeTreeFinder (TTF) tool, which automatically processes millions of articles to discover timetree-containing articles. We estimate that the TTF tool would produce twice as many timetree-containing articles as those discovered manually, whose inclusion in the TT database would potentially double the knowledge accessible to a wider community. Manual inspection showed that the precision on out-of-distribution recently published articles is 87%. This automation will speed up the collection and curation of timetrees with much lower human and time costs. AVAILABILITY AND IMPLEMENTATION: https://github.com/marija-stanojevic/time-tree-classification. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The company we keep. Using hemodialysis social network data to classify patients' kidney transplant attitudes with machine learning algorithms.

BACKGROUND: Hemodialysis clinic patient social networks may reinforce positive and negative attitudes towards kidney transplantation. We examined whether a patient's position within the hemodialysis clinic social network could improve machine learning classification of the patient's positive or negative attitude towards kidney transplantation when compared to sociodemographic and clinical variables. METHODS: We conducted a cross-sectional social network survey of hemodialysis patients in two geographically and demographically different hemodialysis clinics. We evaluated whether machine learning logistic regression models using sociodemographic or network data best predicted the participant's transplant attitude. Models were evaluated for accuracy, precision, recall, and F1-score. RESULTS: The 110 surveyed participants' mean age was 60 ± 13 years old. Half (55%) identified as male, and 74% identified as Black. At facility 1, 69% of participants had a positive attitude towards transplantation whereas at facility 2, 45% of participants had a positive attitude. The machine learning logistic regression model using network data alone obtained a higher accuracy and F1 score than the sociodemographic and clinical data model (accuracy 65% ± 5% vs. 61% ± 7%, F1 score 76% ± 2% vs. 70% ± 7%). A model with a combination of both sociodemographic and network data had a higher accuracy of 74% ± 3%, and an F1-score of 81% ± 2%. CONCLUSION: Social network data improved the machine learning algorithm's ability to classify attitudes towards kidney transplantation, further emphasizing the importance of hemodialysis clinic social networks on attitudes towards transplant.

Deep Learning vs Traditional Models for Predicting Hospital Readmission among Patients with Diabetes.

A hospital readmission risk prediction tool for patients with diabetes based on electronic health record (EHR) data is needed. The optimal modeling approach, however, is unclear. In 2,836,569 encounters of 36,641 diabetes patients, deep learning (DL) long short-term memory (LSTM) models predicting unplanned, all-cause, 30-day readmission were developed and compared to several traditional models. Models used EHR data defined by a Common Data Model. The LSTM model Area Under the Receiver Operating Characteristic Curve (AUROC) was significantly greater than that of the next best traditional model [LSTM 0.79 vs Random Forest (RF) 0.72, p<0.0001]. Experiments showed that performance of the LSTM models increased as prior encounter number increased up to 30 encounters. An LSTM model with 16 selected laboratory tests yielded equivalent performance to a model with all 981 laboratory tests. This new DL model may provide the basis for a more useful readmission risk prediction tool for diabetes patients.

The backbone network of dynamic functional connectivity.

Temporal networks have become increasingly pervasive in many real-world applications, including the functional connectivity analysis of spatially separated regions of the brain. A major challenge in analysis of such networks is the identification of noise confounds, which introduce temporal ties that are nonessential, or links that are formed by chance due to local properties of the nodes. Several approaches have been suggested in the past for static networks or temporal networks with binary weights for extracting significant ties whose likelihood cannot be reduced to the local properties of the nodes. In this work, we propose a data-driven procedure to reveal the irreducible ties in dynamic functional connectivity of resting-state fMRI data with continuous weights. This framework includes a null model that estimates the latent characteristics of the distributions of temporal links through optimization, followed by a statistical test to filter the links whose formation can be reduced to the activities and local properties of their interacting nodes. We demonstrate the benefits of this approach by applying it to a resting-state fMRI dataset, and provide further discussion on various aspects and advantages of it.

Does Whom Patients Sit Next to during Hemodialysis Affect Whether They Request a Living Donation?

Background: The seating arrangement of in-center hemodialysis is conducive to patients forming a relationship and a social network. We examined how seating in the in-center hemodialysis clinic affected patients forming relationships, whether patients formed relationships with others who have similar transplant behaviors (homophily), and whether these relationships influenced patients (social contagion) to request a living donation from family and friends outside of the clinic. Methods: In this 30-month, prospective cohort study, we observed the relationships of 46 patients on hemodialysis in a hemodialysis clinic. Repeated participant surveys assessed in-center transplant discussions and living-donor requests. A separable temporal exponential random graph model estimated how seating, demographics, in-center transplant discussions, and living-donor requests affected relationship formation via sociality and homophily. We examined whether donation requests spread via social contagion using a susceptibility-infected model. Results: For every seat apart, the odds of participants forming a relationship decreased (OR, 0.74; 95% CI, 0.61 to 0.90; P=0.002). Those who requested a living donation tended to form relationships more than those who did not (sociality, OR, 1.6; 95% CI, 1.02 to 2.6; P=0.04). Participants who discussed transplantation in the center were more likely to form a relationship with another participant who discussed transplantation than with someone who did not discuss transplantation (homophily, OR, 1.9; 95% CI, 1.03 to 3.5; P=0.04). Five of the 36 susceptible participants made a request after forming a relationship with another patient. Conclusions: Participants formed relationships with those they sat next to and had similar transplant behaviors. The observed increase in in-center transplant discussions and living-donation requests by the members of the hemodialysis-clinic social network was not because of social contagion. Instead, participants who requested a living donation were more social, formed more relationships within the clinic, and discussed transplantation with each other as a function of health-behavior homophily.

DescribePROT: database of amino acid-level protein structure and function predictions.

We present DescribePROT, the database of predicted amino acid-level descriptors of structure and function of proteins. DescribePROT delivers a comprehensive collection of 13 complementary descriptors predicted using 10 popular and accurate algorithms for 83 complete proteomes that cover key model organisms. The current version includes 7.8 billion predictions for close to 600 million amino acids in 1.4 million proteins. The descriptors encompass sequence conservation, position specific scoring matrix, secondary structure, solvent accessibility, intrinsic disorder, disordered linkers, signal peptides, MoRFs and interactions with proteins, DNA and RNAs. Users can search DescribePROT by the amino acid sequence and the UniProt accession number and entry name. The pre-computed results are made available instantaneously. The predictions can be accesses via an interactive graphical interface that allows simultaneous analysis of multiple descriptors and can be also downloaded in structured formats at the protein, proteome and whole database scale. The putative annotations included by DescriPROT are useful for a broad range of studies, including: investigations of protein function, applied projects focusing on therapeutics and diseases, and in the development of predictors for other protein sequence descriptors. Future releases will expand the coverage of DescribePROT. DescribePROT can be accessed at http://biomine.cs.vcu.edu/servers/DESCRIBEPROT/.

Does Sex, Race, and the Size of a Kidney Transplant Candidate's Social Network Affect the Number of Living Donor Requests? A Multicenter Social Network Analysis of Patients on the Kidney Transplant Waitlist.

BACKGROUND: A kidney transplant candidate's social network serves as a pool of potential living donors. Sex and racial differences in network size, network strength, and living donor requests may contribute to disparities in living donor kidney transplantation. METHODS: In this multicenter cross-sectional study, we performed an egocentric network analysis via a telephone survey of 132 waitlisted candidates (53% female and 69% Black) to identify demographic and network factors associated with requesting living kidney donations. RESULTS: Female participants made requests to more network members than male participants: incidence rate ratio (IRR) 1.95, 95% confidence interval (CI) [1.24-3.06], P < 0.01. Black participants tended to make more requests than whites (IRR 1.65, 95% CI [0.99-2.73], P = 0.05). The number of requests increased with the size of the network (IRR 1.09, 95% CI [1.02-1.16], P = 0.01); however, network size did not differ by sex or race. Network members who provided greater instrumental support to the candidates were most likely to receive a request: odds ratio 1.39, 95% CI [1.08-1.78], P = 0.01. CONCLUSIONS: Transplant candidates' networks vary in size and in the number of requests made to the members. Previously observed racial and sex disparities in living donor kidney transplantation do not appear to be related to network size or to living donation requests, but rather to the network members themselves. Future living donor interventions should focus on the network members and be tailored to their relationship with the candidate.

Influence of medical domain knowledge on deep learning for Alzheimer's disease prediction.

BACKGROUND AND OBJECTIVE: Alzheimer's disease (AD) is the most common type of dementia that can seriously affect a person's ability to perform daily activities. Estimates indicate that AD may rank third as a cause of death for older people, after heart disease and cancer. Identification of individuals at risk for developing AD is imperative for testing therapeutic interventions. The objective of the study was to determine could diagnostics of AD from EMR data alone (without relying on diagnostic imaging) be significantly improved by applying clinical domain knowledge in data preprocessing and positive dataset selection rather than setting naïve filters. METHODS: Data were extracted from the repository of heterogeneous ambulatory EMR data, collected from primary care medical offices all over the U.S. Medical domain knowledge was applied to build a positive dataset from data relevant to AD. Selected Clinically Relevant Positive (SCRP) datasets were used as inputs to a Long-Short-Term Memory (LSTM) Recurrent Neural Network (RNN) deep learning model to predict will the patient develop AD. RESULTS: Risk scores prediction of AD using the drugs domain information in an SCRP AD dataset of 2,324 patients achieved high out-of-sample score - 0.98-0.99 Area Under the Precision-Recall Curve (AUPRC) when using 90% of SCRP dataset for training. AUPRC dropped to 0.89 when training the model using less than 1,500 cases from the SCRP dataset. The model was still significantly better than when using naïve dataset selection. CONCLUSION: The LSTM RNN method that used data relevant to AD performed significantly better when learning from the SCRP dataset than when datasets were selected naïvely. The integration of qualitative medical knowledge for dataset selection and deep learning technology provided a mechanism for significant improvement of AD prediction. Accurate and early prediction of AD is significant in the identification of patients for clinical trials, which can possibly result in the discovery of new drugs for treatments of AD. Also, the contribution of the proposed predictions of AD is a better selection of patients who need imaging diagnostics for differential diagnosis of AD from other degenerative brain disorders.

Predicting complications of diabetes mellitus using advanced machine learning algorithms.

OBJECTIVE: We sought to predict if patients with type 2 diabetes mellitus (DM2) would develop 10 selected complications. Accurate prediction of complications could help with more targeted measures that would prevent or slow down their development. MATERIALS AND METHODS: Experiments were conducted on the Healthcare Cost and Utilization Project State Inpatient Databases of California for the period of 2003 to 2011. Recurrent neural network (RNN) long short-term memory (LSTM) and RNN gated recurrent unit (GRU) deep learning methods were designed and compared with random forest and multilayer perceptron traditional models. Prediction accuracy of selected complications were compared on 3 settings corresponding to minimum number of hospitalizations between diabetes diagnosis and the diagnosis of complications. RESULTS: The diagnosis domain was used for experiments. The best results were achieved with RNN GRU model, followed by RNN LSTM model. The prediction accuracy achieved with RNN GRU model was between 73% (myocardial infarction) and 83% (chronic ischemic heart disease), while accuracy of traditional models was between 66% - 76%. DISCUSSION: The number of hospitalizations was an important factor for the prediction accuracy. Experiments with 4 hospitalizations achieved significantly better accuracy than with 2 hospitalizations. To achieve improved accuracy deep learning models required training on at least 1000 patients and accuracy significantly dropped if training datasets contained 500 patients. The prediction accuracy of complications decreases over time period. Considering individual complications, the best accuracy was achieved on depressive disorder and chronic ischemic heart disease. CONCLUSIONS: The RNN GRU model was the best choice for electronic medical record type of data, based on the achieved results.

Use of disease embedding technique to predict the risk of progression to end-stage renal disease.

The accurate prediction of progression of Chronic Kidney Disease (CKD) to End Stage Renal Disease (ESRD) is of great importance to clinicians and a challenge to researchers as there are many causes and even more comorbidities that are ignored by the traditional prediction models. We examine whether utilizing a novel low-dimensional embedding model disease2disease (D2D) learned from a large-scale electronic health records (EHRs) could well clusters the causes of kidney diseases and comorbidities and further improve prediction of progression of CKD to ESRD compared to traditional risk factors. The study cohort consists of 2,507 hospitalized Stage 3 CKD patients of which 1,375 (54.8%) progressed to ESRD within 3 years. We evaluated the proposed unsupervised learning framework by applying a regularized logistic regression model and a cox proportional hazard model respectively, and compared the accuracies with the ones obtained by four alternative models. The results demonstrate that the learned low-dimensional disease representations from EHRs can capture the relationship between vast arrays of diseases, and can outperform traditional risk factors in a CKD progression prediction model. These results can be used both by clinicians in patient care and researchers to develop new prediction methods.

A heuristic information cluster search approach for precise functional brain mapping.

Detection of the relevant brain regions for characterizing the distinction between cognitive conditions is one of the most sought after objectives in neuroimaging research. A popular approach for achieving this goal is the multivariate pattern analysis which is currently conducted through a number of approaches such as the popular searchlight procedure. This is due to several advantages such as being automatic and flexible with regards to size of the search region. However, these approaches suffer from a number of limitations which can lead to misidentification of truly informative regions which in turn results in imprecise information maps. These limitations mainly stem from several factors such as the fact that the information value of the search spheres are assigned to the voxel at the center of them (in case of searchlight), the requirement for manual tuning of parameters such as searchlight radius and shape, and high complexity and low interpretability in commonly used machine learning-based approaches. Other drawbacks include overlooking the structure and interactions within the regions, and the disadvantages of using certain regularization techniques in analysis of datasets with characteristics of common functional magnetic resonance imaging data. In this article, we propose a fully data-driven maximum relevance minimum redundancy search algorithm for detecting precise information value of the clusters within brain regions while alleviating the above-mentioned limitations. Moreover, in order to make the proposed method faster, we propose an efficient algorithmic implementation. We evaluate and compare the proposed algorithm with the searchlight procedure as well as least absolute shrinkage and selection operator regularization-based mapping approach using both real and synthetic datasets. The analysis results of the proposed approach demonstrate higher information detection precision and map specificity compared to the benchmark approaches.

Time-to-event estimation by re-defining time.

The primary goal of a time-to-event estimation model is to accurately infer the occurrence time of a target event. Most existing studies focus on developing new models to effectively utilize the information in the censored observations. In this paper, we propose a model to tackle the time-to-event estimation problem from a completely different perspective. Our model relaxes a fundamental constraint that the target variable, time, is a univariate number which satisfies a partial order. Instead, the proposed model interprets each event occurrence time as a time concept with a vector representation. We hypothesize that the model will be more accurate and interpretable by capturing (1) the relationships between features and time concept vectors and (2) the relationships among time concept vectors. We also propose a scalable framework to simultaneously learn the model parameters and time concept vectors. Rigorous experiments and analysis have been conducted in medical event prediction task on seven gene expression datasets. The results demonstrate the efficiency and effectiveness of the proposed model. Furthermore, similarity information among time concept vectors helped in identifying time regimes, thus leading to a potential knowledge discovery related to the human cancer considered in our experiments.

Optimizing clinical trials recruitment via deep learning.

OBJECTIVE: Clinical trials, prospective research studies on human participants carried out by a distributed team of clinical investigators, play a crucial role in the development of new treatments in health care. This is a complex and expensive process where investigators aim to enroll volunteers with predetermined characteristics, administer treatment(s), and collect safety and efficacy data. Therefore, choosing top-enrolling investigators is essential for efficient clinical trial execution and is 1 of the primary drivers of drug development cost. MATERIALS AND METHODS: To facilitate clinical trials optimization, we propose DeepMatch (DM), a novel approach that builds on top of advances in deep learning. DM is designed to learn from both investigator and trial-related heterogeneous data sources and rank investigators based on their expected enrollment performance on new clinical trials. RESULTS: Large-scale evaluation conducted on 2618 studies provides evidence that the proposed ranking-based framework improves the current state-of-the-art by up to 19% on ranking investigators and up to 10% on detecting top/bottom performers when recruiting investigators for new clinical trials. DISCUSSION: The extensive experimental section suggests that DM can provide substantial improvement over current industry standards in several regards: (1) the enrollment potential of the investigator list, (2) the time it takes to generate the list, and (3) data-informed decisions about new investigators. CONCLUSION: Due to the great significance of the problem at hand, related research efforts are set to shift the paradigm of how investigators are chosen for clinical trials, thereby optimizing and automating them and reducing the cost of new therapies.

Social network analysis for better understanding of influenza.

INTRODUCTION: The objective of this study is to improve the understanding of spatial spreading of complicated cases of influenza that required hospitalizations, by creating heatmaps and social networks. They will allow to identify critical hubs and routes of spreading of Influenza, in specific geographic locations, in order to contain infections and prevent complications, that require hospitalizations. MATERIAL AND METHODS: Data were downloaded from the Healthcare Cost and Utilization Project (HCUP) - SID, New York State database. Patients hospitalized with flu complications, between 2003 and 2012 were included in the research (30,380 cases). A novel approach was designed, by constructing heatmaps for specific geographic regions in New York state and power law networks, in order to analyze distribution of hospitalized flu cases. RESULTS: Heatmaps revealed that distributions of patients follow urban areas and big roads, indicating that flu spreads along routes, that people use to travel. A scale-free network, created from correlations among zip codes, discovered that, the highest populated zip codes didn't have the largest number of patients with flu complications. Among the top five most affected zip codes, four were in Bronx. Demographics of top affected zip codes were presented in results. Normalized numbers of cases per population revealed that, none of zip codes from Bronx were in the top 20. All zip codes with the highest node degrees were in New York City area. DISCUSSION: Heatmaps identified geographic distribution of hospitalized flu patients and network analysis identified hubs of the infection. Our results will enable better estimation of resources for prevention and treatment of hospitalized patients with complications of Influenza. CONCLUSION: Analyses of geographic distribution of hospitalized patients with Influenza and demographic characteristics of populations, help us to make better planning and management of resources for Influenza patients, that require hospitalization. Obtained results could potentially help to save many lives and improve the health of the population.

Pain-Free Blood Glucose Monitoring Using Wearable Sensors: Recent Advancements and Future Prospects.

Keeping track of blood glucose levels non-invasively is now possible due to diverse breakthroughs in wearable sensors technology coupled with advanced biomedical signal processing. However, each user might have different requirements and priorities when it comes to selecting a self-monitoring solution. After extensive research and careful selection, we have presented a comprehensive survey on noninvasive/pain-free blood glucose monitoring methods from the recent five years (2012-2016). Several techniques, from bioinformatics, computer science, chemical engineering, microwave technology, etc., are discussed in order to cover a wide variety of solutions available for different scales and preferences. We categorize the noninvasive techniques into nonsample- and sample-based techniques, which we further grouped into optical, nonoptical, intermittent, and continuous. The devices manufactured or being manufactured for noninvasive monitoring are also compared in this paper. These techniques are then analyzed based on certain constraints, which include time efficiency, comfort, cost, portability, power consumption, etc., a user might experience. Recalibration, time, and power efficiency are the biggest challenges that require further research in order to satisfy a large number of users. In order to solve these challenges, artificial intelligence (AI) has been employed by many researchers. AI-based estimation and decision models hold the future of noninvasive glucose monitoring in terms of accuracy, cost effectiveness, portability, efficiency, etc. The significance of this paper is twofold: first, to bridge the gap between IT and medical field; and second, to bridge the gap between end users and the solutions (hardware and software).