Showing papers by "Ernest Nlandu Kamavuako published in 2022"

The Effect of Signal Duration on the Classification of Heart Sounds: A Deep Learning Approach

Abstract: Deep learning techniques are the future trend for designing heart sound classification methods, making conventional heart sound segmentation dispensable. However, despite using fixed signal duration for training, no study has assessed its effect on the final performance in detail. Therefore, this study aims at analysing the duration effect on the commonly used deep learning methods to provide insight for future studies in data processing, classifier, and feature selection. The results of this study revealed that (1) very short heart sound signal duration (1 s) weakens the performance of Recurrent Neural Networks (RNNs), whereas no apparent decrease in the tested Convolutional Neural Network (CNN) model was found. (2) RNN outperformed CNN using Mel-frequency cepstrum coefficients (MFCCs) as features. There was no difference between RNN models (LSTM, BiLSTM, GRU, or BiGRU). (3) Adding dynamic information (∆ and ∆²MFCCs) of the heart sound as a feature did not improve the RNNs’ performance, and the improvement on CNN was also minimal (≤2.5% in MAcc). The findings provided a theoretical basis for further heart sound classification using deep learning techniques when selecting the input length.

The Effect of EMG Features on the Classification of Swallowing Events and the Estimation of Fluid Intake Volume

Abstract: Nowadays, society is experiencing an increase in the number of adults aged 65 and over, and it is projected that the older adult population will triple in the coming decades. As older adults are prone to becoming dehydrated, which can significantly impact healthcare costs and staff, it is necessary to advance healthcare technologies to cater to such needs. However, there has not been an extensive research effort to implement a device that can autonomously track fluid intake. In particular, the ability of surface electromyographic sensors (sEMG) to monitor fluid intake has not been investigated in depth. Our previous study demonstrated a reasonable classification and estimation ability of sEMG using four features. This study aimed to examine if classification and estimation could be potentiated by combining an optimal subset of features from a library of forty-six time and frequency-domain features extracted from the data recorded using eleven subjects. Results demonstrated a classification accuracy of 95.94 ± 2.76% and an f-score of 94.93 ± 3.51% in differentiating between liquid swallows from non-liquid swallowing events using five features only, and a volume estimation RMSE of 2.80 ± 1.22 mL per sip and an average estimation error of 15.43 ± 8.64% using two features only. These results are encouraging and prove that sEMG could be a potential candidate for monitoring fluid intake.

Paroxysmal Atrial Fibrillation Detection by Combined Recurrent Neural Network and Feature Extraction on ECG Signals

Abstract: Paroxysmal atrial fibrillation (AFib) or intermittent atrial fibrillation is one type of atrial fibrillation which occurs rapidly and stops spontaneously within days. Its episodes can last several seconds, hours, or even days before returning to normal sinus rhythm. A lack of intervention may lead the paroxysmal into persistent atrial fibrillation, causing severe risk to human health. However, due to its intermittent characteristics, it is generally neglected by patients. Therefore, real-time monitoring and accurate automatic algorithms are highly needed for early screening. This study proposes a two-stage algorithm, including a BiLSTM network to classify healthy and atrial fibrillation, followed by a feature-extraction-based neural network (NN) to identify the persistent, paroxysmal atrial fibrillation onsets. The extracted features include the entropy and standard deviation of the RR intervals. The two steps can achieve 90.14% and 92.56% accuracy in the validation sets on small segments. This overall algorithm also has the advantage of the low computing load, which shows a high potential for a portable embedded device.

Scalable tensor factorization for recovering multiday missing intramuscular electromyography data

Abstract: To design a prosthetic hand which can classify movements based on the electromyography (EMG) signals, complete and good quality signals are essential. However, due to different reasons such as disconnection of electrodes or muscles fatigue the recorded EMG data can be incomplete, which degrades the classification of test movements. In this paper, we first acquire multiday intramuscular EMG (iEMG) signals (which are invasive) with higher Signal-to-Noise Ratio (SNR) compared to surface EMG (sEMG) signals; followed by application of matrix (non-negative matrix factorization – NMF) and tensor factorization methods (Canonical Polyadic Decomposition (CPD), Tucker decomposition (TD) & Canonical Polyadic-Weighted Optimization (CP-WOPT)) for recovering structured missing data i.e., chunks of missing samples in channels. Furthermore, we tested the scalability of NMF, CPD, TD and CP-WOPT by employing them on the large multiday (seven days) iEMG data where the size of missing data is increased from day 1 to day 7, and for each day a fixed percentage of missing data is introduced from 10% to worst case of 50%. Results show that CP-WOPT outperformed NMF, CPD and TD to recover large percentage of missing data in terms of Relative Mean Error (RME) even when 7 days of data is considered. CP-WOPT showed robustness even for the worse case even when 50% iEMG data is removed from day 1 to day 7 where it’s RME degraded slightly from 0.08 to 0.1.

Hierarchical Multi-Scale Convolutional Network for Murmurs Detection on PCG Signals

Abstract: Computer-aided analysis is helpful in improving heart sound classification. PhysioNet Challenge 2022 provides a platform for researchers to evaluate their proposed classification algorithms. In the Challenge, our team (HearTech+) proposed a recording quality assessment method based on frequency density distribution for label correction to prevent the poor-quality recording segments from misleading network optimisation. Besides, a hierarchical multi-scale convolutional neural network (HMS-Net) was designed to conduct both the murmur $(T1)$ and clinical outcome $(T2)$ classification. HMS-Net extracts convolutional features from the spectrograms on multiple scales and fuses them through its hierarchical architecture. The network builds long short-term independencies between multi-scale features and improves the classification performance. Finally, the prediction of a patient is based on the ensembled segment predictions by sliding window. In the five-fold cross-validation by patients, the proposed algorithm performed an average weighted accuracy of 0.81 (best 0.853) on $T1$ and an average challenge score of 9808 (best 9242) on $T2$. In the Challenge hidden validation set, the proposed algorithm achieved 0.806 weighted accuracy on $T1$ and 9120 challenge score on $T2$, ranking 1st and $4^{th}$ out of 305 entries, respectively. In the final hidden testing set, $T1$ was 0.776 ranking $2^{nd}$, and $T2$ was 12069.

Spectral Image-Based Multiday Surface Electromyography Classification of Hand Motions Using CNN for Human–Computer Interaction

Abstract: Physiological signals such as electromyography (EMG) have been used in human–computer interaction (HCI) for medical applications. Wearable prostheses, such as robotic limbs, have seen a surge in popularity because of technological advancements in myoelectric interfaces. In spite of encouraging achievements with pattern-recognition-based control systems, user acceptability of prosthetic hands still needs improvement in control robustness. The purpose of this research is to compare multiday surface EMG (sEMG) recordings and measure the performance of convolutional neural network (CNN) to enhance myoelectric control. The performance metrics used in this study are accuracy, macro weighted precision (MWP), macro weighted recall (MWR), and macro ${F}1$ -score for eight able-bodied (healthy and nondisabled) and four amputee subjects. Using Mel-spectrogram-based sEMG data from both the able-bodied (healthy and nondisabled) and amputee participants, our proposed CNN has achieved a mean classification accuracy of 99.42% ± 0.42% and 98.00% ± 0.58% for the able-bodied (healthy and nondisabled) and amputee subjects for the within-day analysis, respectively. The proposed CNN outperformed other classifiers ( ${p} \leq0.05$ ) in the between-day analysis for twofold (65.88% ± 10.1% and 58.37% ± 9.11%) and for sevenfold validation (88.73% ± 1.43% and 77.35% ± 2.72%) using sEMG recordings from the able-bodied (healthy and nondisabled) and amputee subjects, respectively. The proposed CNN is compared with the pretrained transfer learning (TL) models and has achieved higher accuracy with lower computational cost. The results demonstrate that CNN can considerably increase the effectiveness of the pattern recognition myoelectric control schemes and can extract deep information from EMG data.

Time-Frequency Distributions of Heart Sound Signals: A Comparative Study using Convolutional Neural Networks

Abstract: Time-Frequency Distributions (TFDs) support the heart sound characterisation and classification in early cardiac screening. However, despite the frequent use of TFDs in signal analysis, no study comprehensively compared their performances on deep learning for automatic diagnosis. Furthermore, the combination of signal processing methods as inputs for Convolutional Neural Networks (CNNs) has been proved as a practical approach to increasing signal classification performance. Therefore, this study aimed to investigate the optimal use of TFD/ combined TFDs as input for CNNs. The presented results revealed that: 1) The transformation of the heart sound signal into the TF domain achieves higher classification performance than using of raw signals. Among the TFDs, the difference in the performance was slight for all the CNN models (within $1.3\%$ in average accuracy). However, Continuous wavelet transform (CWT) and Chirplet transform (CT) outperformed the rest. 2) The appropriate increase of the CNN capacity and architecture optimisation can improve the performance, while the network architecture should not be overly complicated. Based on the ResNet or SEResNet family results, the increase in the number of parameters and the depth of the structure do not improve the performance apparently. 3) Combining TFDs as CNN inputs did not significantly improve the classification results. The findings of this study provided the knowledge for selecting TFDs as CNN input and designing CNN architecture for heart sound classification.

Surface EMG Statistical and Performance Analysis of Targeted-Muscle-Reinnervated (TMR) Transhumeral Prosthesis Users in Home and Laboratory Settings

Abstract: A pattern-recognition (PR)-based myoelectric control system is the trend of future prostheses development. Compared with conventional prosthetic control systems, PR-based control systems provide high dexterity, with many studies achieving >95% accuracy in the last two decades. However, most research studies have been conducted in the laboratory. There is limited research investigating how EMG signals are acquired when users operate PR-based systems in their home and community environments. This study compares the statistical properties of surface electromyography (sEMG) signals used to calibrate prostheses and quantifies the quality of calibration sEMG data through separability indices, repeatability indices, and correlation coefficients in home and laboratory settings. The results demonstrate no significant differences in classification performance between home and laboratory environments in within-calibration classification error (home: 6.33 ± 2.13%, laboratory: 7.57 ± 3.44%). However, between-calibration classification errors (home: 40.61 ± 9.19%, laboratory: 44.98 ± 12.15%) were statistically different. Furthermore, the difference in all statistical properties of sEMG signals is significant (p < 0.05). Separability indices reveal that motion classes are more diverse in the home setting. In summary, differences in sEMG signals generated between home and laboratory only affect between-calibration performance.

On the Applications of EMG Sensors and Signals

Abstract: The ability to execute limb motions derives from composite command signals (or efferent signals) that stem from the central nervous system through the highway of the spinal cord and peripheral nerves to the muscles that drive the joints [...].