scholarly journals Blind Source Separation for the Aggregation of Machine Learning Algorithms: An Arrhythmia Classification Case

Electronics ◽  
2020 ◽  
Vol 9 (3) ◽  
pp. 425
Author(s):  
Krzysztof Gajowniczek ◽  
Iga Grzegorczyk ◽  
Michał Gostkowski ◽  
Tomasz Ząbkowski
Keyword(s):  
Blind Source Separation ◽  
Cardiac Arrhythmias ◽  
Classification Accuracy ◽  
Recursive Partitioning ◽  
Source Separation ◽  
Classification Problem ◽  
Machine Learning Algorithms ◽  
Physiological Signals ◽  
New Approach ◽  
Model Aggregation

In this work, we present an application of the blind source separation (BSS) algorithm to reduce false arrhythmia alarms and to improve the classification accuracy of artificial neural networks (ANNs). The research was focused on a new approach for model aggregation to deal with arrhythmia types that are difficult to predict. The data for analysis consisted of five-minute-long physiological signals (ECG, BP, and PLETH) registered for patients with cardiac arrhythmias. For each patient, the arrhythmia alarm occurred at the end of the signal. The data present a classification problem of whether the alarm is a true one—requiring attention or is false—should not have been generated. It was confirmed that BSS ANNs are able to detect four arrhythmias—asystole, ventricular tachycardia, ventricular fibrillation, and tachycardia—with higher classification accuracy than the benchmarking models, including the ANN, random forest, and recursive partitioning and regression trees. The overall challenge scores were between 63.2 and 90.7.

scholarly journals A New Criterion for Bounded Component Analysis

2021 ◽  
Author(s):  
Renan Brotto ◽  
Kenji Nose-Filho ◽  
João M. T. Romano
Keyword(s):  
Blind Source Separation ◽  
Source Separation ◽  
Component Analysis ◽  
Rotation Matrix ◽  
Separation Problem ◽  
Correlated Sources ◽  
New Approach ◽  
Bounded Component ◽  
New Criterion

<div>In this paper we present a new criterion for bounded component analysis, a quite new approach for the Blind Source Separation problem. For the determined case, we show that the `1-norm of the estimated sources can be used as a contrast for the problem. We present a blind algorithm for the source separation of independents sources or mixtures of correlated sources by only a rotation matrix. We also present a variety of simulations assessing the performance of the proposed approach.</div>

scholarly journals Noncontact Heart Rate Measurement Using a Webcam, Based on Joint Blind Source Separation and a Skin Reflection Model: For a Wide Range of Imaging Conditions

2021 ◽  
Vol 2021 ◽  
pp. 1-18
Author(s):  
Boyuan Zhang ◽  
Hengkang Li ◽  
Lisheng Xu ◽  
Lin Qi ◽  
Yudong Yao ◽  
...  
Keyword(s):  
Heart Rate ◽  
Blind Source Separation ◽  
Body Movement ◽  
Source Separation ◽  
New Approach ◽  
Factors Affecting ◽  
Reflection Model ◽  
Wide Range ◽  
Rate Information ◽  
Imaging Conditions

Remote photoplethysmography (rPPG) can be used for noncontact and continuous measurement of the heart rate (HR). Currently, the main factors affecting the accuracy and robustness of rPPG-based HR measurement methods are the subject’s skin tone, body movement, exercise recovery, and variable or inadequate illumination. In response to these challenges, this study is aimed at investigating a rPPG-based HR measurement method that is effective under a wide range of conditions by only using a webcam. We propose a new approach, which combines joint blind source separation (JBSS) and a projection process based on a skin reflection model, so as to eliminate the interference of background illumination and enhance the extraction of pulse rate information. Three datasets derived from subjects with different skin tones considering six environmental scenarios are used to validate the proposed method against three other state-of-the-art methods. The results show that the proposed method can provide more accurate and robust HR measurement for all three datasets and is therefore more applicable to a wide range of scenarios.

Cognitive Biometrics

2011 ◽  
pp. 105-136
Author(s):  
Kenneth Revett
Keyword(s):  
Classification Accuracy ◽  
User Authentication ◽  
Single Point ◽  
Machine Learning Algorithms ◽  
States Of Mind ◽  
Continuous Mode ◽  
Static Mode ◽  
New Approach ◽  
The Individual ◽  
Unique Signature

Cognitive biometrics is a new authentication scheme that utilises the cognitive, emotional, and conative state of an individual as the basis of user authentication and/or identification. These states of mind (and their derivatives) are extracted by recording various biosignals such as the EEG, ECG, and electrodermal response (EDR) of the individual in response to the presentation of the authentication stimulus. Stimuli are selected which elicit characteristic changes within the acquired biosignal(s) that represent unique responses from the individual. These characteristic changes are processed using a variety of machine learning algorithms, resulting in a unique signature that identifies or authenticates the individual. This approach can be applied in both static mode (single point of authentication), or in continuous mode, either alone, or in a multi-modal approach. The data suggest that the classification accuracy can reach 100% in many scenarios, providing support for the efficacy of this new approach to both static and continuous biometrics.

scholarly journals Blind Source Separation In The Analysis Of Electrocardiogram Pre-Shock Waveforms During Ventricular Fibrillation

2021 ◽  
Author(s):  
Marzieh Rasooli
Keyword(s):  
Ventricular Fibrillation ◽  
Blind Source Separation ◽  
Treatment Option ◽  
Electric Shock ◽  
Classification Accuracy ◽  
Source Separation ◽  
Temporal Phase ◽  
Spatio Temporal ◽  
The Given ◽  
Extract Information

Ventricular fibrillation (VF) is a lethal cardiac arrhythmia and electric shock is the only available treatment option for it. Existing works focus on predicting shock success to help improve cardiac resuscitation outcomes. It is desirable to extract information from the electrograms that relates to the current theories on VF mechanism and associate them to the prediction of shock outcomes. To this effect this study used a unique human VF database to evaluate the independent sources (ISs) extracted from Blind Source Separation approach (BSS) and a correlation of 88% was observed between the dominant ISs extracted using a single lead ECG with the number of rotors (i.e., sources identified using multi-channel spatio-temporal phase maps) supporting the hypothesis that the ISs are associated with the rotors. In predicting the shock outcomes using features extracted from the ISs for the given database, we achieved a classification accuracy of 68%.

Letter to the Editor: RIVAL PENALIZED COMPETITIVE LEARNING BASED APPROACH FOR DISCRETE-VALUED SOURCE SEPARATION

2000 ◽  
Vol 10 (06) ◽  
pp. 483-490 ◽  
Author(s):  
YIU-MING CHEUNG ◽  
LEI XU
Keyword(s):  
Blind Source Separation ◽  
Source Separation ◽  
Competitive Learning ◽  
Cluster Number ◽  
Number Of Clusters ◽  
New Approach ◽  
Noise Interference ◽  
Correct Number ◽  
Mixing Matrix ◽  
Source Number

This paper presents an approach based on Rival Penalized Competitive Learning (RPCL) rules for discrete-valued source separation. In this approach, we first build a connection between the source number and the cluster number of observations. Then, we use the RPCL rule to automatically find out the correct number of clusters such that the source number is determined. Moreover, we tune the de-mixing matrix based on the cluster centers instead of the observation themselves, whereby the noise interference is considerably reduced. The experiments have shown that this new approach not only quickly and automatically determines the number of sources, but also is insensitive to the noise in performing blind source separation.

Sign in / Sign up

Export Citation Format

Share Document