One-dimensional convolutional neural network architecture for classification of mental tasks from electroencephalogram

Saini, Manali; Satija, Udit; Upadhayay, Madhur Deo

doi:10.1016/j.bspc.2022.103494

Cited by 28 publications

(13 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To evaluate our proposed framework, we have used the following performance metrics: accuracy (ACC), precision (PRC), sensitivity (SEN)/recall (REC), specificity (SPE), and Mathew's correlation coefficient (MCC) for both binary and multiclass classification similar to [49], [50]. These metrics can be computed from the confusion matrix [51].…”

Section: A Performance Metricsmentioning

confidence: 99%

A Novel Melspectrogram Snippet Representation Learning Framework for Severity Detection of Chronic Obstructive Pulmonary Diseases

Roy

Satija

2023

IEEE Trans. Instrum. Meas.

Self Cite

View full text Add to dashboard Cite

Chronic obstructive pulmonary disease (COPD) is a major public health concern across the world. Since it is an incurable disease, early detection and accurate diagnosis are very crucial for preventing the progression of the disease. Lung sounds provide reliable and accurate prognoses for identifying respiratory diseases. Recently, Altan et al. recorded 12-channel real-time lung sound dataset, namely RespiratoryDatabase@TR, for five different severity levels of COPD at Antakya State Hospital Turkey, and proposed deep learning frameworks for two-class COPD classification and five-class classification using a deep belief network (DBN) classifier and extreme learning machine (ELM) classifier respectively. A classification accuracy of 95.84% and 94.31% were achieved for two-class and five-class respectively. In this paper, we have proposed a melspectrogram snippet representation learning framework for both two-class and five-class COPD classification. The proposed framework consists of the following stages: preprocessing, melspectrogram snippet representation generation from lung sound and fine tuning of a pretrained YAMNet. Experimental analysis on the RespiratoryDatabase@TR dataset demonstrates that the proposed framework achieves accuracies of 99.25% and 96.14% for binary and multi-class COPD severity classification respectively, which is superior to the only existing methods proposed by Altan et al. for severity analysis of COPD using lung sounds.

show abstract

Section: A Performance Metricsmentioning

confidence: 99%

A Novel Melspectrogram Snippet Representation Learning Framework for Severity Detection of Chronic Obstructive Pulmonary Diseases

Roy

Satija

2023

IEEE Trans. Instrum. Meas.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Additionally, to the best of our knowledge, this is the second work on asthma classification based on lung sounds following to Altan et al [13]. To evaluate the performance of the proposed RDsLINet, we have employed the following performance metrics: accuracy (Acc), specificity (Spe), recall (Rec) or sensitivity (Sen), and precision (Prc) to evaluate our proposed framework similar to [28], [29], [30], [31]. The performance parameters can be computed from confusion matrix [32].…”

Section: A Classification Tasks and Performance Metricsmentioning

confidence: 99%

RDsLINet: A Novel Lightweight Inception Network For Respiratory Disease Classification Using Lung Sounds

Roy¹,

Satija²

2022

Preprint

View full text Add to dashboard Cite

<p>Objective: Respiratory diseases are the world’s third leading cause of mortality. Early detection is critical in dealing with respiratory diseases as it improves the effectiveness of intervention, including treatment and reducing the spread. The main aim of this paper is to propose a novel lightweight inception network to classify wide spectrum of respiratory diseases using lung sound signals. Methods: The proposed framework consists of three stages: (a) preprocessing, (b) melspectrogram extraction and conversion into a 3-channel image, lastly (c) classification of the melspectrogram images into different pathological classes using proposed lightweight inception network, namely RDsLINet. Results: Utilizing the proposed architecture, we have achieved high classification accuracy of 96%, 99.5%, 91.6% for seven class classification, six class classification and healthy vs. asthma classification. To the best of our knowledge this is the first work on seven class respiratory disease classification using lung sounds. Whereas, our proposed network outperforms all the existing published works for six class and binary classification. Conclusion: The suggested framework makes use of deep learning methods and offers a standardized evaluation with strong categorization capabilities. In order to distinguish between a wide range of respiratory diseases, our study is a pioneering one that focuses exclusively on lung sounds. Significance: The proposed framework can be translated to real time clinical application which will facilitate the prospect of automated respiratory health screening using lung sounds.</p>

show abstract

“…Thanks to this advantage, it is frequently preferred in studies [11]. With CNN, one-dimensional non-image data such as signals and sounds can be processed, and it has been successfully applied in the analysis of EEG signals [12,18]. CNN has a structure that can consist of several layers.…”

Section: Proposed 1d Cnn Modelmentioning

confidence: 99%

“…A few studies evaluating MWL using machine learning classifiers have been mentioned above. Recently, it has been noted that higher success rates can be obtained from raw data by using the automatic feature extraction advantage of CNN [11,12]. Although CNN is frequently used in BCI studies, it has been emphasized that there is a lack of work in MWL-level decomposed and classification [12].…”

Section: Introductionmentioning

confidence: 99%

“…Recently, it has been noted that higher success rates can be obtained from raw data by using the automatic feature extraction advantage of CNN [11,12]. Although CNN is frequently used in BCI studies, it has been emphasized that there is a lack of work in MWL-level decomposed and classification [12]. In this study, it is aimed to distinguish the 1D-CNN classifier and MWL (L-H) by decomposed the EEG signals recorded during the activity into subbands using the EMD method.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Classification Mental Workload Levels from EEG Signals with 1D Convolutional Neural Network

Baydemir

Latifoğlu

Orhanbulucu

2022

EJRnD

View full text Add to dashboard Cite

Mental workload (MWL) can be estimated according to the state of cognitive capacity after an activity. In this study, it is aimed to classify MWL levels from Electroencephalogram (EEG) signals recorded from a task moment. Using the proposed one-dimensional convolutional neural network (1D-CNN) model in the study, low (L) and high (H) level WL states were classified. The classification process was carried out in two stages. EEG signals passed through the preprocessing stage were classified with 1D-CNN in the first stage. In the second step, these signals were decomposed into subbands by applying Empirical Mode Decomposition (EMD) and classified with 1D-CNN. As a result of the classification process, accuracy (Acc), sensitivity (Sens), and specificity (Spe) values were obtained and evaluated in this study. As a result of the evaluation, the most successful Acc rate was 98.4%, Sens rate 97.62%, and Spe rate 98.94%

show abstract

One-dimensional convolutional neural network architecture for classification of mental tasks from electroencephalogram

Cited by 28 publications

References 41 publications

A Novel Melspectrogram Snippet Representation Learning Framework for Severity Detection of Chronic Obstructive Pulmonary Diseases

A Novel Melspectrogram Snippet Representation Learning Framework for Severity Detection of Chronic Obstructive Pulmonary Diseases

RDsLINet: A Novel Lightweight Inception Network For Respiratory Disease Classification Using Lung Sounds

Classification Mental Workload Levels from EEG Signals with 1D Convolutional Neural Network

Contact Info

Product

Resources

About