Reconfigurable Architecture for Multi-lead ECG Signal Compression with High-frequency Noise Reduction

Chowdhury, Mehdi Hasan; Cheung, Ray C. C.

doi:10.1038/s41598-019-53460-3

Cited by 19 publications

(12 citation statements)

References 46 publications

(47 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Which require time and efforts [7] . Transformation domain methods transform ECG signal to frequency domain or other domains to compress ECG signal, then an inverse transformation is performed to reconstruct the original signal [14][15][16][17][18] . Parameter extraction method is based on extracted discriminated features from original signal to reduce the size of ECG signal [4] .…”

Section: Discussion and Resultsmentioning

confidence: 99%

“…Direct methods require time and efforts in designing an appropriate compressing algorithm signals through the extraction of a subset of samples from the original signal sample set, they are usually fast because they use heuristic analysis [7] . Transformation domain methods transform ECG signal to frequency domain or other domains to analyze energy distribution and compress ECG signal, then an inverse transformation is performed to reconstruct the original signal [14][15][16][17][18] . Parameter extraction method is based on extracted optimized and discriminated features from original signal to reduce the size of ECG signal [4] .…”

Section: Compression Methods For Ecg Signalsmentioning

confidence: 99%

See 1 more Smart Citation

Bidirectional Recurrent Nets for ECG Signal Compression

AL-Saidi

Hindi

2022

j. of Comput. sci. res.

View full text Add to dashboard Cite

Electrocardiogram (ECG) is a commonly used tool in biological diagnosis of heart diseases. ECG allows the representation of electrical signals which cause heart muscles to contract and relax. Recently, accurate deep learning methods have been developed to overcome manual diagnosis in terms of time and effort. However, most of current automatic medical diagnosis use long electrocardiogram (ECG) signals to inspect different types of heart arrhythmia. Therefore, ECG signal files tend to require large storage to store and may cause significant overhead when exchanged over a computer network. This raises the need to come up with effective compression methods for ECG signals. In this work, the authors investigate using BERT (Bidirectional Encoder Representations from Transformers) model, which is a bidirectional neural network that was originally designed for natural language. The authors evaluate the model with respect to its compression ratio and information preservation, and measure information preservation in terms of the of the accuracy of a convolutional neural network in classifying the decompressed signal. The results show that the method can achieve up to 83% saving in storage. Also, the classification accuracy of the decompressed signals is around 92.41%. Furthermore, the method enables the user to balance the compression ratio and the required accuracy of the CNN classifiers.

show abstract

Section: Discussion and Resultsmentioning

confidence: 99%

Section: Compression Methods For Ecg Signalsmentioning

confidence: 99%

Bidirectional Recurrent Nets for ECG Signal Compression

AL-Saidi

Hindi

2022

j. of Comput. sci. res.

View full text Add to dashboard Cite

show abstract

“…It is a crucial component of a common e-health system, as ECG signals must frequently be compacted for lengthy data storage and distant delivery. High-speed parallel computes units, especially the field-programmable gate array (FPGA), and flexible software features are available with reconfigurable architecture [ 31 , 32 ].…”

Section: Methodsmentioning

confidence: 99%

Reconfigurable Architectures with High-Frequency Noise Suppression for Wearable ECG Devices

Jerard

Thilagaraj

Pandiaraj

et al. 2021

Journal of Healthcare Engineering

View full text Add to dashboard Cite

Recent advances in electronics and microelectronics have aided the development of low-cost devices that are widely used as well-being or preventive monitoring devices by many people. Remote health monitoring, which includes wearable sensors, actuators, and modern communication and information systems, offers effective programs that allow people to live peacefully in their own homes while also being protected in some way. High-frequency noise, power-line interface, and baseline drift are prevalent during the data-acquisition system of an ECG signal, and they can limit signal understanding. They (noises) must be isolated in order to provide an appropriate diagnostic of the patient. When removing high-frequency components (noise) from an ECG signal with an FIR filter, the critical path delay increases considerably as the filter's duration increases. To reduce high-frequency noise, simple moving average filters with pipelining and look-ahead transformation techniques are extensively used in this study. With the use of pipelining and look-ahead techniques, the only objective is to increase the clock speed of the designs. The moving average filters (conventional and proposed) were created on an Altera Cyclone IV FPGA EP4CE115F29C7 chip using the Quartus II software v13.1 tool. Finally, performance metrics such logic elements, clock speed, and power consumption were compared and studied thoroughly. The recursive pipelined 8-tap MA filter with look-ahead approach outperforms the other designs (685.48 MHz) in this investigation.

show abstract

“…A blood pressure measure out of this range may be associated with the incidence of several cardiovascular events (e.g., stroke, heart failure, and end-stage renal disease) [ 6 , 47 ]. Since arterial pressure varies continuously during the cardiac cycle, the morphological shape of different signals associated with blood pressure varies as well, as seen in Figure 2 c. This shape is acquired as an electrocardiograph (ECG), ballistocardiograph (BCG), and phonocardiograph (PCG) signals [ 48 , 49 , 50 ]. Also, the morphological shapes and values of blood pressure wave vary when it travels from the highly elastic central arteries to the stiffer ones [ 51 ], and when it travels away from the heart either upward (i.e., towards the head) or downward (i.e., towards the foot) in an upright position ( Figure 3 ) [ 43 , 47 ].…”

Section: Blood Pressure Measurementmentioning

confidence: 99%

Blood Pressure Sensors: Materials, Fabrication Methods, Performance Evaluations and Future Perspectives

Al-Qatatsheh

Morsi

Zavabeti

et al. 2020

Sensors

View full text Add to dashboard Cite

Advancements in materials science and fabrication techniques have contributed to the significant growing attention to a wide variety of sensors for digital healthcare. While the progress in this area is tremendously impressive, few wearable sensors with the capability of real-time blood pressure monitoring are approved for clinical use. One of the key obstacles in the further development of wearable sensors for medical applications is the lack of comprehensive technical evaluation of sensor materials against the expected clinical performance. Here, we present an extensive review and critical analysis of various materials applied in the design and fabrication of wearable sensors. In our unique transdisciplinary approach, we studied the fundamentals of blood pressure and examined its measuring modalities while focusing on their clinical use and sensing principles to identify material functionalities. Then, we carefully reviewed various categories of functional materials utilized in sensor building blocks allowing for comparative analysis of the performance of a wide range of materials throughout the sensor operational-life cycle. Not only this provides essential data to enhance the materials’ properties and optimize their performance, but also, it highlights new perspectives and provides suggestions to develop the next generation pressure sensors for clinical use.

show abstract

Reconfigurable Architecture for Multi-lead ECG Signal Compression with High-frequency Noise Reduction

Cited by 19 publications

References 46 publications

Bidirectional Recurrent Nets for ECG Signal Compression

Bidirectional Recurrent Nets for ECG Signal Compression

Reconfigurable Architectures with High-Frequency Noise Suppression for Wearable ECG Devices

Blood Pressure Sensors: Materials, Fabrication Methods, Performance Evaluations and Future Perspectives

Contact Info

Product

Resources

About