scholarly journals Early Sepsis Detection with Deep Learning on EHR Event Sequences

2019 ◽  
Vol 2 (3) ◽  
pp. 39
Author(s):  
Simon Meyer Lauritsen ◽  
Mads Ellersgaard Kalør ◽  
Emil Lund Kongsgaard ◽  
Bo Thiesson
Keyword(s):  
Deep Learning ◽  
Domain Knowledge ◽  
Healthcare Professionals ◽  
Short Term Memory ◽  
Sequence Data ◽  
Positive Outcome ◽  
Learning System ◽  
High Morbidity ◽  
Event Sequences ◽  
Sepsis Detection

Background: Sepsis is a clinical condition involving an extreme inflammatory response to an infection, and is associated with high morbidity and mortality. Without intervention, this response can progress to septic shock, organ failure and death. Every hour that treatment is delayed mortality increases. Early identification of sepsis is therefore important for a positive outcome. Methods: We constructed predictive models for sepsis detection and performed a register-based cohort study on patients from four Danish municipalities. We used event-sequences of raw electronic health record (EHR) data from 2013 to 2017, where each event consists of three elements: a timestamp, an event category (e.g. medication code), and a value. In total, we consider 25.622 positive (SIRS criteria) sequences and 25.622 negative sequences with a total of 112 million events distributed across 64 different hospital units. The number of potential predictor variables in raw EHR data easily exceeds 10.000 and can be challenging for predictive modeling due to this large volume of sparse, heterogeneous events. Traditional approaches have dealt with this complexity by curating a limited number of variables of importance; a labor-intensive process that may discard a vast majority of information. In contrast, we consider a deep learning system constructed as a combination of a convolutional neural network (CNN) and long short-term memory (LSTM) network. Importantly, our system learns representations of the key factors and interactions from the raw event sequence data itself. Results: Our model predicts sepsis with an AUROC score of 0.8678, at 11 hours before actual treatment was started, outperforming all currently deployed approaches. At other prediction times, the model yields following AUROC scores. 15 min: 0.9058, 3 hours: 0.8803, 24 hours: 0.8073. Conclusion: We have presented a novel approach for early detection of sepsis that has more true positives and fewer false negatives than existing alarm systems without introducing domain knowledge into the model. Importantly, the model does not require changes in the daily workflow of healthcare professionals at hospitals, as the model is based on data that is routinely captured in the EHR. This also enables real-time prediction, as healthcare professionals enters the raw events in the EHR.

Abstract 279: Multi-task Learning Improves Model Performance in Predicting Rare Catastrophic Events in Healthcare Claims Dataset

Circulation ◽  
2020 ◽  
Vol 142 (Suppl_4) ◽  
Author(s):  
ChienYu Chi ◽  
Yen-Pin Chen ◽  
Adrian Winkler ◽  
Kuan-Chun Fu ◽  
Fie Xu ◽  
...  
Keyword(s):  
Cardiac Arrest ◽  
Deep Learning ◽  
Short Term Memory ◽  
Learning System ◽  
Research Database ◽  
Learning Models ◽  
Catastrophic Events ◽  
Single Task ◽  
Task Learning ◽  
Hospital Cardiac Arrest

Introduction: Predicting rare catastrophic events is challenging due to lack of targets. Here we employed a multi-task learning method and demonstrated that substantial gains in accuracy and generalizability was achieved by sharing representations between related tasks Methods: Starting from Taiwan National Health Insurance Research Database, we selected adult people (>20 year) experienced in-hospital cardiac arrest but not out-of-hospital cardiac arrest during 8 years (2003-2010), and built a dataset using de-identified claims of Emergency Department (ED) and hospitalization. Final dataset had 169,287 patients, randomly split into 3 sections, train 70%, validation 15%, and test 15%.Two outcomes, 30-day readmission and 30-day mortality are chosen. We constructed the deep learning system in two steps. We first used a taxonomy mapping system Text2Node to generate a distributed representation for each concept. We then applied a multilevel hierarchical model based on long short-term memory (LSTM) architecture. Multi-task models used gradient similarity to prioritize the desired task over auxiliary tasks. Single-task models were trained for each desired task. All models share the same architecture and are trained with the same input data Results: Each model was optimized to maximize AUROC on the validation set with the final metrics calculated on the held-out test set. We demonstrated multi-task deep learning models outperform single task deep learning models on both tasks. While readmission had roughly 30% positives and showed miniscule improvements, the mortality task saw more improvement between models. We hypothesize that this is a result of the data imbalance, mortality occurred roughly 5% positive; the auxiliary tasks help the model interpret the data and generalize better. Conclusion: Multi-task deep learning models outperform single task deep learning models in predicting 30-day readmission and mortality in in-hospital cardiac arrest patients.

scholarly journals EEG and Deep Learning Based Brain Cognitive Function Classification

Computers ◽  
2020 ◽  
Vol 9 (4) ◽  
pp. 104
Author(s):  
Saraswati Sridhar ◽  
Vidya Manian
Keyword(s):  
Deep Learning ◽  
Motor Imagery ◽  
Cognitive Aging ◽  
Classification Accuracy ◽  
Short Term Memory ◽  
Principal Component ◽  
Learning System ◽  
Detection Accuracy ◽  
Band Power ◽  
Signal Features

Electroencephalogram signals are used to assess neurodegenerative diseases and develop sophisticated brain machine interfaces for rehabilitation and gaming. Most of the applications use only motor imagery or evoked potentials. Here, a deep learning network based on a sensory motor paradigm (auditory, olfactory, movement, and motor-imagery) that employs a subject-agnostic Bidirectional Long Short-Term Memory (BLSTM) Network is developed to assess cognitive functions and identify its relationship with brain signal features, which is hypothesized to consistently indicate cognitive decline. Testing occurred with healthy subjects of age 20–40, 40–60, and >60, and mildly cognitive impaired subjects. Auditory and olfactory stimuli were presented to the subjects and the subjects imagined and conducted movement of each arm during which Electroencephalogram (EEG)/Electromyogram (EMG) signals were recorded. A deep BLSTM Neural Network is trained with Principal Component features from evoked signals and assesses their corresponding pathways. Wavelet analysis is used to decompose evoked signals and calculate the band power of component frequency bands. This deep learning system performs better than conventional deep neural networks in detecting MCI. Most features studied peaked at the age range 40–60 and were lower for the MCI group than for any other group tested. Detection accuracy of left-hand motor imagery signals best indicated cognitive aging (p = 0.0012); here, the mean classification accuracy per age group declined from 91.93% to 81.64%, and is 69.53% for MCI subjects. Motor-imagery-evoked band power, particularly in gamma bands, best indicated (p = 0.007) cognitive aging. Although the classification accuracy of the potentials effectively distinguished cognitive aging from MCI (p < 0.05), followed by gamma-band power.

scholarly journals A clinical deep learning framework for continually learning from cardiac signals across diseases, time, modalities, and institutions

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Dani Kiyasseh ◽  
Tingting Zhu ◽  
David Clifton
Keyword(s):  
Deep Learning ◽  
Domain Knowledge ◽  
Learning Strategy ◽  
Learning System ◽  
Destructive Interference ◽  
Diagnostic Systems ◽  
Time Data ◽  
Data Points ◽  
Physiological Sensors ◽  
Continual Learning

AbstractDeep learning algorithms trained on instances that violate the assumption of being independent and identically distributed (i.i.d.) are known to experience destructive interference, a phenomenon characterized by a degradation in performance. Such a violation, however, is ubiquitous in clinical settings where data are streamed temporally from different clinical sites and from a multitude of physiological sensors. To mitigate this interference, we propose a continual learning strategy, entitled CLOPS, that employs a replay buffer. To guide the storage of instances into the buffer, we propose end-to-end trainable parameters, termed task-instance parameters, that quantify the difficulty with which data points are classified by a deep-learning system. We validate the interpretation of these parameters via clinical domain knowledge. To replay instances from the buffer, we exploit uncertainty-based acquisition functions. In three of the four continual learning scenarios, reflecting transitions across diseases, time, data modalities, and healthcare institutions, we show that CLOPS outperforms the state-of-the-art methods, GEM1 and MIR2. We also conduct extensive ablation studies to demonstrate the necessity of the various components of our proposed strategy. Our framework has the potential to pave the way for diagnostic systems that remain robust over time.

scholarly journals Epileptic Seizure Detection using Deep Learning Approach

2019 ◽  
Vol 3 (2) ◽  
pp. 41 ◽  
Author(s):  
Sirwan Tofiq Jaafar ◽  
Mokhtar Mohammadi
Keyword(s):  
Deep Learning ◽  
Epileptic Seizure ◽  
Domain Knowledge ◽  
Short Term Memory ◽  
Time Frequency ◽  
Standard Tool ◽  
Machine Learning Applications ◽  
Memory Network ◽  
Eeg Recordings ◽  
Almost All

An epileptic seizure is a sign of abnormal activity in the human brain. Electroencephalogram (EEG) is a standard tool that has been used vastly for detection of seizure activities. Many methods have been developed to help the neurophysiologists to detect the seizure activities with high accuracy. Most of them rely on the features extracted in the time, frequency, or time-frequency domains. The performance of the proposed methods is related to the performance of the features extracted from EEG recordings. Deep neural networks enable learning directly on the data without the domain knowledge needed to construct a feature set. This approach has been hugely successful in almost all machine learning applications. We propose a new framework that also learns directly from the data, without extracting a feature set. We proposed an original deep-learning-based method to classify EEG recordings. The EEG signal is segmented into 4 s segments and used to train the long- and short-term memory network. The trained model is used to discriminate the EEG seizure from the background. The Freiburg EEG dataset is used to assess the performance of the classifier. The 5-fold cross-validation is selected for evaluating the performance of the proposed method. About 97.75% of the accuracy is achieved.

scholarly journals Classifying COVID-19 variants based on genetic sequences using deep learning models

2021 ◽  
Author(s):  
Sayantani Basu ◽  
Roy H. Campbell
Keyword(s):  
Deep Learning ◽  
Short Term Memory ◽  
Sequence Data ◽  
Virus Disease ◽  
Class Imbalance ◽  
Fixed Number ◽  
Class Imbalance Problem ◽  
Imbalance Problem ◽  
Batch Sizes ◽  
The One

The COrona VIrus Disease (COVID-19) pandemic led to the occurrence of several variants with time. This has led to an increased importance of understanding sequence data related to COVID-19. In this chapter, we propose an alignment-free k-mer based LSTM (Long Short-Term Memory) deep learning model that can classify 20 different variants of COVID-19. We handle the class imbalance problem by sampling a fixed number of sequences for each class label. We handle the vanishing gradient problem in LSTMs arising from long sequences by dividing the sequence into fixed lengths and obtaining results on individual runs. Our results show that one- vs-all classifiers have test accuracies as high as 92.5% with tuned hyperparameters compared to the multi-class classifier model. Our experiments show higher overall accuracies for B.1.1.214, B.1.177.21, B.1.1.7, B.1.526, and P.1 on the one-vs-all classifiers, suggesting the presence of distinct mutations in these variants. Our results show that embedding vector size and batch sizes have insignificant improvement in accuracies, but changing from 2-mers to 3-mers mostly improves accuracies. We also studied individual runs which show that most accuracies improved after the 20th run, indicating that these sequence positions may have more contributions to distinguishing among different COVID-19 variants.

scholarly journals Design and development of learning model for compression and processing of deoxyribonucleic acid genome sequence

2022 ◽  
Vol 12 (2) ◽  
pp. 1786
Author(s):  
Raveendra Gudodagi ◽  
Rayapur Venkata Siva Reddy ◽  
Mohammed Riyaz Ahmed
Keyword(s):  
Deep Learning ◽  
Data Compression ◽  
Data Storage ◽  
Genome Sequence ◽  
Sequence Data ◽  
Genomic Data ◽  
Modern Technology ◽  
Optimization Techniques ◽  
Learning System ◽  
Systematic Analysis

Owing to the substantial volume of human genome sequence data files (from 30-200 GB exposed) Genomic data compression has received considerable traction and storage costs are one of the major problems faced by genomics laboratories. This involves a modern technology of data compression that reduces not only the storage but also the reliability of the operation. There were few attempts to solve this problem independently of both hardware and software. A systematic analysis of associations between genes provides techniques for the recognition of operative connections among genes and their respective yields, as well as understandings into essential biological events that are most important for knowing health and disease phenotypes. This research proposes a reliable and efficient deep learning system for learning embedded projections to combine gene interactions and gene expression in prediction comparison of deep embeddings to strong baselines. In this paper we preform data processing operations and predict gene function, along with gene ontology reconstruction and predict the gene interaction. The three major steps of genomic data compression are extraction of data, storage of data, and retrieval of the data. Hence, we propose a deep learning based on computational optimization techniques which will be efficient in all the three stages of data compression.

scholarly journals Development of a Deep Learning System To Reduce The Time Needed for COVID-19 RT-PCR: Can We Use Deep Learning To Get RT-PCR Test Results of COVID-19 Faster?

2021 ◽  
Author(s):  
YOONJE LEE ◽  
Yu-Seop KIM ◽  
Da-in Lee ◽  
Seri Jeong ◽  
Gu-Hyun Kang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Sensitivity And Specificity ◽  
Diagnostic Performance ◽  
Operating Characteristic ◽  
Short Term Memory ◽  
Learning System ◽  
Test Results ◽  
Rt Pcr ◽  
Pcr Diagnosis ◽  
Long Short Term Memory

Abstract Reducing the time to diagnose COVID-19 helps to manage insufficient isolation-bed resources and adequately accommodate critically ill patients in clinical fields. There is currently no alternative method to RT-PCR, which requires 40 cycles to diagnose COVID-19. We proposed a deep learning (DL) model to improve the speed of COVID-19 RT-PCR diagnosis. We developed and tested a DL model using the long-short term memory method with a dataset of fluorescence values measured in each cycle of 5,810 RT-PCR tests. Among the DL models developed here, the diagnostic performance of the 21st model showed an area under the receiver operating characteristic (AUROC), sensitivity, and specificity of 84.55%, 93.33%, and 75.72%, respectively. The diagnostic performance of the 24th model showed an AUROC sensitivity, and specificity of 91.27%, 90.00%, and 92.54%, respectively.

scholarly journals Large-scale comparative assessment of computational predictors for lysine post-translational modification sites

2018 ◽  
Vol 20 (6) ◽  
pp. 2267-2290 ◽  
Author(s):  
Zhen Chen ◽  
Xuhan Liu ◽  
Fuyi Li ◽  
Chen Li ◽  
Tatiana Marquez-Lago ◽  
...  
Keyword(s):  
Feature Selection ◽  
Deep Learning ◽  
Large Scale ◽  
Short Term Memory ◽  
Sequence Data ◽  
Model Performance ◽  
Structural Features ◽  
Computational Techniques ◽  
Post Translational Modification ◽  
Sequencing Data

Abstract Lysine post-translational modifications (PTMs) play a crucial role in regulating diverse functions and biological processes of proteins. However, because of the large volumes of sequencing data generated from genome-sequencing projects, systematic identification of different types of lysine PTM substrates and PTM sites in the entire proteome remains a major challenge. In recent years, a number of computational methods for lysine PTM identification have been developed. These methods show high diversity in their core algorithms, features extracted and feature selection techniques and evaluation strategies. There is therefore an urgent need to revisit these methods and summarize their methodologies, to improve and further develop computational techniques to identify and characterize lysine PTMs from the large amounts of sequence data. With this goal in mind, we first provide a comprehensive survey on a large collection of 49 state-of-the-art approaches for lysine PTM prediction. We cover a variety of important aspects that are crucial for the development of successful predictors, including operating algorithms, sequence and structural features, feature selection, model performance evaluation and software utility. We further provide our thoughts on potential strategies to improve the model performance. Second, in order to examine the feasibility of using deep learning for lysine PTM prediction, we propose a novel computational framework, termed MUscADEL (Multiple Scalable Accurate Deep Learner for lysine PTMs), using deep, bidirectional, long short-term memory recurrent neural networks for accurate and systematic mapping of eight major types of lysine PTMs in the human and mouse proteomes. Extensive benchmarking tests show that MUscADEL outperforms current methods for lysine PTM characterization, demonstrating the potential and power of deep learning techniques in protein PTM prediction. The web server of MUscADEL, together with all the data sets assembled in this study, is freely available at http://muscadel.erc.monash.edu/. We anticipate this comprehensive review and the application of deep learning will provide practical guide and useful insights into PTM prediction and inspire future bioinformatics studies in the related fields.

A Deep Learning based Arabic Script Recognition System: Benchmark on KHAT

2020 ◽  
Vol 17 (3) ◽  
pp. 299-305 ◽  
Author(s):  
Riaz Ahmad ◽  
Saeeda Naz ◽  
Muhammad Afzal ◽  
Sheikh Rashid ◽  
Marcus Liwicki ◽  
...  
Keyword(s):  
Deep Learning ◽  
Character Recognition ◽  
Data Augmentation ◽  
Short Term Memory ◽  
Recognition System ◽  
Learning Approach ◽  
Arabic Text ◽  
Data Set ◽  
Processing Step ◽  
Handwritten Arabic

This paper presents a deep learning benchmark on a complex dataset known as KFUPM Handwritten Arabic TexT (KHATT). The KHATT data-set consists of complex patterns of handwritten Arabic text-lines. This paper contributes mainly in three aspects i.e., (1) pre-processing, (2) deep learning based approach, and (3) data-augmentation. The pre-processing step includes pruning of white extra spaces plus de-skewing the skewed text-lines. We deploy a deep learning approach based on Multi-Dimensional Long Short-Term Memory (MDLSTM) networks and Connectionist Temporal Classification (CTC). The MDLSTM has the advantage of scanning the Arabic text-lines in all directions (horizontal and vertical) to cover dots, diacritics, strokes and fine inflammation. The data-augmentation with a deep learning approach proves to achieve better and promising improvement in results by gaining 80.02% Character Recognition (CR) over 75.08% as baseline.

Human Activity Recognition using Fourier Transform Inspired Deep Learning Combination Model

2019 ◽  
Vol 9 (1) ◽  
pp. 16-31
Author(s):  
Kyungkoo Jun
Keyword(s):  
Fourier Transform ◽  
Deep Learning ◽  
Short Term Memory ◽  
Window Size ◽  
Sensor Data ◽  
Data Sets ◽  
Data Set ◽  
Proposed Model ◽  
Testing Data ◽  
Labeling Scheme

Background & Objective: This paper proposes a Fourier transform inspired method to classify human activities from time series sensor data. Methods: Our method begins by decomposing 1D input signal into 2D patterns, which is motivated by the Fourier conversion. The decomposition is helped by Long Short-Term Memory (LSTM) which captures the temporal dependency from the signal and then produces encoded sequences. The sequences, once arranged into the 2D array, can represent the fingerprints of the signals. The benefit of such transformation is that we can exploit the recent advances of the deep learning models for the image classification such as Convolutional Neural Network (CNN). Results: The proposed model, as a result, is the combination of LSTM and CNN. We evaluate the model over two data sets. For the first data set, which is more standardized than the other, our model outperforms previous works or at least equal. In the case of the second data set, we devise the schemes to generate training and testing data by changing the parameters of the window size, the sliding size, and the labeling scheme. Conclusion: The evaluation results show that the accuracy is over 95% for some cases. We also analyze the effect of the parameters on the performance.

Sign in / Sign up

Export Citation Format

Share Document