Anomaly detection in ECG time signals via deep long short-term memory networks

By Sucheta Chauhan, Lovekesh Vig. International Conference on Data Science and Advanced Analytics, 2015

391 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cardiology and Cardiovascular Medicine, Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/54c35ae24a4d6a48ef691baf17f1441a17909181

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Cardiology and Cardiovascular Medicine

In Control and Systems Engineering

Citations in the corpus

DCdetector: Dual Attention Contrastive Representation Learning for Time Series Anomaly Detection (2023, background): “Most of the supervised and semisupervised methods [14, 18, 51, 53, 87, 90] can not handle the challenge of limited labeled data, especially the anomalies are dynamic and new anomalies never observed before may occur.”
Deep Learning for Unsupervised Insider Threat Detection in Structured Cybersecurity Data Streams (2017, neutral): “Early work on modeling normal user activity on a network using RNNs was performed by Debar et al. (1992).”
Converging blockchain and next-generation artificial intelligence technologies to decentralize and accelerate biomedical research and healthcare (2015, background): “Anomalies in the data could also be linked to health conditions, so both those approaches could be used for pathology detection in health recordings [103, 104].”
Anomaly Detection for IoT Time-Series Data: A Survey (2020, methodology): “A similar deep LSTM network has been applied to ECG signals in [72] to identify a variety of different anomalous signals, again using the multiple time-step ahead probabilistic error measure.”
LSTM-based Encoder-Decoder for Multi-sensor Anomaly Detection (2016, methodology): “Recently, deep LSTMs have been used as prediction models in LSTM-AD (Malhotra et al., 2015; Chauhan & Vig, 2015; Yadav et al.) where a prediction model learnt over the normal time-series using LSTM networks is used to predict future points, and likelihood of prediction error is used as a measure of…”
Electrocardiogram generation with a bidirectional LSTM-CNN generative adversarial network (2019, methodology): “LSTM has been applied to tasks based on time series data such as anomaly detection in ECG signals 27 .”
Multi-head CNN-RNN for multi-time series anomaly detection: An industrial case study (2019, methodology): “Recent approaches use DL algorithms such as Long-Short Term Memory (LSTM) to model the normal behavior of a sequence to create a predictive model and then calculate the anomaly score of an observation as the deviation from the predicted value [43, 44, 45, 46, 47, 48].”
Deep Learning for Anomaly Detection: A Survey (2019, background): “A survey of traditional anomaly detection techniques and its challenges to detect anomalies in social networks is a well studied topic in literature Liu and Chawla [2017], Savage et al.”
DeepAnT: A Deep Learning Approach for Unsupervised Anomaly Detection in Time Series (2019, methodology): “Available: https:// goo.gl/CPKYrc [6] P. Malhotra, L. Vig, G. Shroff, and P. Agarwal, ‘‘Long short term memory networks for anomaly detection in time series,’’ in Proc.”
Unsupervised Anomaly Detection With LSTM Neural Networks (2017, methodology): “Here, they allocate a different set of sequences to learn the parameters of the distribution and threshold via the maximum likelihood estimation technique [10], [25].”
Smart Anomaly Detection in Sensor Systems (2020, methodology): “[62] proposed using LSTM to capture anomalies in Electrocardiography (ECG) signals.”
LSTM-Based ECG Classification for Continuous Monitoring on Personal Wearable Devices (2018, background): “For instance, many only consider a selected subset of the standard classes [26]–[31], which makes the design and training of neural networks much simpler because not all the challenging cases are included.”
Deep Learning for Time Series Anomaly Detection: A Survey (2022, neutral): “Thus, it would be very beneficial to adopt an anomaly detection approach for analysing ECG signals which are developed in [96], [196] and [28].”
Automated Atrial Fibrillation Detection using a Hybrid CNN-LSTM Network on Imbalanced ECG Datasets (2021, background): “data, but also can selectively remember or forget information checking the current memory state – and that made them popular for ECG signal classification [18,21].”
Anomaly Detection in Time Series: A Comprehensive Evaluation (2022, methodology): “DeepAnT [94] STORN [123] CoalESN [99] DeepNAP [72] DeepLSTM[31] Hybrid KNN [124] EncDec-AD [88] LSTM-based”