scholarly journals Automatic Classification of Myocardial Infarction Using Spline Representation of Single-Lead Derived Vectorcardiography

Sensors ◽  
2020 ◽  
Vol 20 (24) ◽  
pp. 7246
Author(s):  
Yu-Hung Chuang ◽  
Chia-Ling Huang ◽  
Wen-Whei Chang ◽  
Jen-Tzung Chien
Keyword(s):  
Myocardial Infarction ◽  
Short Term Memory ◽  
Spline Approximation ◽  
Three Dimensional ◽  
Patient Specific ◽  
Reconstruction Method ◽  
Ecg Signals ◽  
Multilayer Perceptron Network ◽  
Lstm Network ◽  
Induced Changes

Myocardial infarction (MI) is one of the most prevalent cardiovascular diseases worldwide and most patients suffer from MI without awareness. Therefore, early diagnosis and timely treatment are crucial to guarantee the life safety of MI patients. Most wearable monitoring devices only provide single-lead electrocardiography (ECG), which represents a major limitation for their applicability in diagnosis of MI. Incorporating the derived vectorcardiography (VCG) techniques can help monitor the three-dimensional electrical activities of human hearts. This study presents a patient-specific reconstruction method based on long short-term memory (LSTM) network to exploit both intra- and inter-lead correlations of ECG signals. MI-induced changes in the morphological and temporal wave features are extracted from the derived VCG using spline approximation. After the feature extraction, a classifier based on multilayer perceptron network is used for MI classification. Experiments on PTB diagnostic database demonstrate that the proposed system achieved satisfactory performance to differentiating MI patients from healthy subjects and to localizing the infarcted area.

scholarly journals HPCWMF: A Hybrid Predictive Cloud Workload Management Framework Using Improved LSTM Neural Network

2020 ◽  
Vol 20 (4) ◽  
pp. 55-73
Author(s):  
K. Dinesh Kumar ◽  
E. Umamaheswari
Keyword(s):  
Short Term Memory ◽  
Hybrid Approach ◽  
Differential Evolution Algorithm ◽  
Memory Loss ◽  
Data Sets ◽  
Workload Prediction ◽  
Workload Management ◽  
Management Framework ◽  
Multilayer Perceptron Network ◽  
Lstm Network

AbstractFor cloud providers, workload prediction is a challenging task due to irregular incoming workloads from users. Accurate workload prediction is essential for scheduling the resources to the cloud applications. Thus, in this paper, the authors propose a predictive cloud workload management framework to estimate the needed resources in advance based on a hybrid approach, which is a combination of an improved Long Short-Term Memory (LSTM) network and a multilayer perceptron network. By improving the traditional LSTM architecture by using opposition-based differential evolution algorithm and dropout technique on recurrent connection without memory loss, the proposed approach has the ability to perform a better prediction process. A novel hybrid predictive approach is aiming at enhancing the prediction performance of the cloud workload. Finally, the authors measure the proposed approach’s effectiveness under benchmark data sets of NASA and Saskatchewan servers. The experimental results proved that the proposed approach outperforms the other conventional methods.

scholarly journals A Multilayer LSTM Auto-Encoder for Fetal ECG Anomaly Detection

2021 ◽  
Author(s):  
Inna Skarga-Bandurova ◽  
Tetiana Biloborodova ◽  
Illia Skarha-Bandurov ◽  
Yehor Boltov ◽  
Maryna Derkach
Keyword(s):  
Anomaly Detection ◽  
Short Term Memory ◽  
Traditional Approach ◽  
Signal Identification ◽  
Term Memory ◽  
Ecg Signals ◽  
Fetal Ecg ◽  
Long Short Term Memory ◽  
Lstm Network ◽  
Filtering Procedure

The paper introduces a multilayer long short-term memory (LSTM) based auto-encoder network to spot abnormalities in fetal ECG. The LSTM network was used to detect patterns in the time series, reconstruct errors and classify a given segment as an anomaly or not. The proposed anomaly detection method provides a filtering procedure able to reproduce ECG variability based on the semi-supervised paradigm. Experiments show that the proposed method can learn better features than the traditional approach without any prior knowledge and subject to proper signal identification can facilitate the analysis of fetal ECG signals in daily life.

scholarly journals An Effective LSTM Recurrent Network to Detect Arrhythmia on Imbalanced ECG Dataset

2019 ◽  
Vol 2019 ◽  
pp. 1-10 ◽  
Author(s):  
Junli Gao ◽  
Hongpo Zhang ◽  
Peng Lu ◽  
Zongmin Wang
Keyword(s):  
Short Term Memory ◽  
Recurrent Network ◽  
Short Term ◽  
Ecg Signals ◽  
Beat Classification ◽  
Long Short Term Memory ◽  
Lstm Network ◽  
Reliable Solution ◽  
Electrocardiogram Ecg

To reduce the high mortality rate from cardiovascular disease (CVD), the electrocardiogram (ECG) beat plays a significant role in computer-aided arrhythmia diagnosis systems. However, the complex variations and imbalance of ECG beats make this a challenging issue. Since ECG beat data exist in heavily imbalanced category, an effective long short-term memory (LSTM) recurrence network model with focal loss (FL) is proposed. For this purpose, the LSTM network can disentangle the timing features in complex ECG signals, while the FL is used to resolve the category imbalance by downweighting easily identified normal ECG examples. The advantages of the proposed network have been verified in the MIT-BIH arrhythmia database. Experimental results show that the LSTM network with FL achieved a reliable solution to the problem of imbalanced datasets in ECG beat classification and was not sensitive to quality of ECG signals. The proposed method can be deployed in telemedicine scenarios to assist cardiologists into more accurately and objectively diagnosing ECG signals.

scholarly journals Spatial-Spectral Network for Hyperspectral Image Classification: A 3-D CNN and Bi-LSTM Framework

2021 ◽  
Vol 13 (12) ◽  
pp. 2353
Author(s):  
Junru Yin ◽  
Changsheng Qi ◽  
Qiqiang Chen ◽  
Jiantao Qu
Keyword(s):  
Short Term Memory ◽  
Hyperspectral Image ◽  
Sequence Data ◽  
Three Dimensional ◽  
Spectral Feature ◽  
Spectral Features ◽  
Spectral Bands ◽  
Current State ◽  
Feature Extractor ◽  
Lstm Network

Recently, deep learning methods based on the combination of spatial and spectral features have been successfully applied in hyperspectral image (HSI) classification. To improve the utilization of the spatial and spectral information from the HSI, this paper proposes a unified network framework using a three-dimensional convolutional neural network (3-D CNN) and a band grouping-based bidirectional long short-term memory (Bi-LSTM) network for HSI classification. In the framework, extracting spectral features is regarded as a procedure of processing sequence data, and the Bi-LSTM network acts as the spectral feature extractor of the unified network to fully exploit the close relationships between spectral bands. The 3-D CNN has a unique advantage in processing the 3-D data; therefore, it is used as the spatial-spectral feature extractor in this unified network. Finally, in order to optimize the parameters of both feature extractors simultaneously, the Bi-LSTM and 3-D CNN share a loss function to form a unified network. To evaluate the performance of the proposed framework, three datasets were tested for HSI classification. The results demonstrate that the performance of the proposed method is better than the current state-of-the-art HSI classification methods.

scholarly journals Application of Heartbeat-Attention Mechanism for Detection of Myocardial Infarction Using 12-Lead ECG Records

2019 ◽  
Vol 9 (16) ◽  
pp. 3328 ◽  
Author(s):  
Zhang ◽  
Li
Keyword(s):  
Myocardial Infarction ◽  
Short Term Memory ◽  
Attention Mechanism ◽  
Short Term ◽  
Sliding Windows ◽  
Long Short Term Memory ◽  
The Difference ◽  
Lstm Network ◽  
Electrocardiogram Ecg ◽  
Ecg Database

Early detection and effective treatment of myocardial infarction can prevent the deterioration of ischemic heart disease and greatly reduce the possibility of sudden death. On the basis of standard 12-lead electrocardiogram (ECG) records, this paper proposes a bidirectional, long short-term memory (Bi-LSTM) network with a heartbeat-attention mechanism to effectively and automatically detect myocardial infarction (MI). First, we divide the standard 12-lead ECG records into sliding windows with the same number of heartbeats. Subsequently, we do not use any labels of heartbeats to train the Bi-LSTM network and the heartbeat-attention mechanism is applied to automatically weight the difference between unlabeled heartbeats. Finally, our method is validated by patients’ complete ECG records and real labels in the Physikalisch-Technische Bundesanstalt (PTB) diagnostic ECG database. When compared with the same network without the heartbeat-attention mechanism or other existing methods, our method achieves comparable or better performance. The accuracy, sensitivity, and specificity reach 94.77%, 95.58%, and 90.48%, respectively.

scholarly journals Myocardial Infarction Classification Based on Convolutional Neural Network and Recurrent Neural Network

2019 ◽  
Vol 9 (9) ◽  
pp. 1879 ◽  
Author(s):  
Kai Feng ◽  
Xitian Pi ◽  
Hongying Liu ◽  
Kai Sun
Keyword(s):  
Neural Network ◽  
Myocardial Infarction ◽  
Convolutional Neural Network ◽  
Short Term Memory ◽  
Rapid Development ◽  
Classification Performance ◽  
Human Beings ◽  
Ecg Signals ◽  
Good Classification Performance ◽  
Memory Network

Myocardial infarction is one of the most threatening cardiovascular diseases for human beings. With the rapid development of wearable devices and portable electrocardiogram (ECG) medical devices, it is possible and conceivable to detect and monitor myocardial infarction ECG signals in time. This paper proposed a multi-channel automatic classification algorithm combining a 16-layer convolutional neural network (CNN) and long-short term memory network (LSTM) for I-lead myocardial infarction ECG. The algorithm preprocessed the raw data to first extract the heartbeat segments; then it was trained in the multi-channel CNN and LSTM to automatically learn the acquired features and complete the myocardial infarction ECG classification. We utilized the Physikalisch-Technische Bundesanstalt (PTB) database for algorithm verification, and obtained an accuracy rate of 95.4%, a sensitivity of 98.2%, a specificity of 86.5%, and an F1 score of 96.8%, indicating that the model can achieve good classification performance without complex handcrafted features.

Three-dimensional reconstruction of the 40S ribosomal subunit from rabbit reticulocytes

1987 ◽  
Vol 45 ◽  
pp. 936-937
Author(s):  
Neng-Yu Zhang ◽  
Terence Wagenknecht ◽  
Michael Radermacher ◽  
Tom Obrig ◽  
Joachim Frank
Keyword(s):  
Three Dimensional ◽  
Ribosomal Subunit ◽  
Uranyl Acetate ◽  
Reconstruction Method ◽  
Single Exposure ◽  
Electron Dose ◽  
Ribosomal Subunits ◽  
40S Ribosomal Subunit ◽  
Dimensional Reconstruction ◽  
Rabbit Reticulocytes

We have reconstructed the 40S ribosomal subunit at a resolution of 4 nm using the single-exposure pseudo-conical reconstruction method of Radermacher et al.Small (40S) ribosomal subunits were Isolated from rabbit reticulocytes, applied to grids and negatively stained (0.5% uranyl acetate) in a manner that “sandwiches” the specimen between two layers of carbon. Regions of the grid exhibiting uniform and thick staining were identified and photographed twice (magnification 49,000X). The first micrograph was always taken with the specimen tilted by 50° and the second was of the Identical area untilted (Fig. 1). For each of the micrographs the specimen was subjected to an electron dose of 2000-3000 el/nm2.Three hundred thirty particles appearing in the L view (defined in [4]) were selected from both tilted- and untilted-specimen micrographs. The untilted particles were aligned and their rotational alignment produced the azimuthal angles of the tilted particles in the conical tilt series.

scholarly journals Numerical computation of blood flow for a patient-specific hemodialysis shunt model

2021 ◽  
Author(s):  
Surabhi Rathore ◽  
Tomoki Uda ◽  
Viet Q. H. Huynh ◽  
Hiroshi Suito ◽  
Toshitaka Watanabe ◽  
...  
Keyword(s):  
Stokes Equations ◽  
Three Dimensional ◽  
Arteriovenous Shunt ◽  
Navier Stokes ◽  
Patient Specific ◽  
Stabilized Finite Element ◽  
Computed Tomography Scanning ◽  
Outflow Boundary ◽  
Stage Renal Disease ◽  
End Stage

AbstractHemodialysis procedure is usually advisable for end-stage renal disease patients. This study is aimed at computational investigation of hemodynamical characteristics in three-dimensional arteriovenous shunt for hemodialysis, for which computed tomography scanning and phase-contrast magnetic resonance imaging are used. Several hemodynamical characteristics are presented and discussed depending on the patient-specific morphology and flow conditions including regurgitating flow from the distal artery caused by the construction of the arteriovenous shunt. A simple backflow prevention technique at an outflow boundary is presented, with stabilized finite element approaches for incompressible Navier–Stokes equations.

scholarly journals Application of subject-specific adaptive mechanical loading for bone healing in a mouse tail vertebral defect

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Angad Malhotra ◽  
Matthias Walle ◽  
Graeme R. Paul ◽  
Gisela A. Kuhn ◽  
Ralph Müller
Keyword(s):  
Mechanical Loading ◽  
Bone Defect ◽  
Bone Healing ◽  
Three Dimensional ◽  
Patient Specific ◽  
Dependent Manner ◽  
Micro Computed Tomography ◽  
Computed Tomography Images ◽  
Bone Defect Healing ◽  
Subject Specific

AbstractMethods to repair bone defects arising from trauma, resection, or disease, continue to be sought after. Cyclic mechanical loading is well established to influence bone (re)modelling activity, in which bone formation and resorption are correlated to micro-scale strain. Based on this, the application of mechanical stimulation across a bone defect could improve healing. However, if ignoring the mechanical integrity of defected bone, loading regimes have a high potential to either cause damage or be ineffective. This study explores real-time finite element (rtFE) methods that use three-dimensional structural analyses from micro-computed tomography images to estimate effective peak cyclic loads in a subject-specific and time-dependent manner. It demonstrates the concept in a cyclically loaded mouse caudal vertebral bone defect model. Using rtFE analysis combined with adaptive mechanical loading, mouse bone healing was significantly improved over non-loaded controls, with no incidence of vertebral fractures. Such rtFE-driven adaptive loading regimes demonstrated here could be relevant to clinical bone defect healing scenarios, where mechanical loading can become patient-specific and more efficacious. This is achieved by accounting for initial bone defect conditions and spatio-temporal healing, both being factors that are always unique to the patient.

Sign in / Sign up

Export Citation Format

Share Document