Biomedical Signal Proceesing and Biomedical Instrumentation
9
Scopus Publications
132
Scholar Citations
6
Scholar h-index
5
Scholar i10-index
Scopus Publications
A Machine Learning Approach to the Smartwatch-based Epileptic Seizure Detection System Gaurav G, Rahul Shukla, Gagandeep Singh, Ashish Kumar Sahani IETE Journal of Research, 2024 Worldwide around 70 million people have epilepsy, and every year, more than 1 out of 1000 cases of epilepsy result in Sudden Unexpected Death in Epilepsy (SUDEP). Video – EEG is the standard clinical method for monitoring epilepsy and seizures. However, wearable systems are required to monitor epileptic activity in daily living due to the complexity of using EEG outside the laboratory. Also, to prevent SUDEP, early prediction of seizure onset is required. In this work, we propose a machine learning model to detect ictal and preictal conditions using an Empatica E4 smartwatch. The Empatica E4 records real-time photoplethysmography, electrodermal activity, accelerometry, and temperature. Clinical data were recorded from 11 patients with epilepsy (PWE) for 19 seizure onsets. Features from all the modalities were extracted by taking segments of the signal during the seizure (ictal), pre-seizure, and inter-ictal (non-seizure) conditions. These features were used on support vector machine (SVM-RBF), decision tree (DTC), and logistic regression (LRC)-based supervised training for ictal vs. non-ictal and pre-ictal vs. inter-ictal conditions. The highest accuracy of 99.40% was recorded for DTC-based seizure detection classifier during 10-fold cross-validation. Also, the highest accuracy of 95.42% was recorded for DTC-based pre-seizure onset detection classifier during 10-fold cross-validation.
Epileptic seizure detection using STFT based peak mean feature and support vector machine Nitin Sharma, Gaurav G, R. S. Anand Proceedings of the 8th International Conference on Signal Processing and Integrated Networks Spin 2021, 2021 Epilepsy is a neurological condition of intermittent brain dysfunction arising from irregular neuronal discharge through the brain. The electroencephalogram (EEG) offers valuable information about the brain’s physiological states and is also an effective method for detecting epilepsy. This study aims to develop a computer-aided automation system to identify epileptic seizures through EEG data from epileptic and healthy subjects. We employed discrete Short-time Fourier transform (STFT) to decompose EEG data into sub-bands, and sample entropy, mean, and peak mean features were extracted from each sub-band. Feature ’mean’ accounts for baseline differences, ’sample entropy’ for the chaotic nature of EEG data, and ’peak mean’ for the amplitude differences between healthy and epileptic EEG data. We achieved the highest classification accuracy of 100% in distinguishing epileptic ictal EEG signals and EEG signals from healthy subjects through 10-fold cross-validation using the Support vector machine with radial basis function (SVM-RBF) classifier. We also presented the comparison of peak mean feature with other well-known features in epilepsy detection using EEG. The high accuracy results obtained by the peak mean feature show its potential in seizure detection using EEG.
Correlating visual and auditory perception and attention level using EEG parameters G. Gaurav, R. S. Anand, Vinod Kumar 2020 7th International Conference on Signal Processing and Integrated Networks Spin 2020, 2020 Focus, alertness level, and reaction time are key cognitive factors and are dependent on visual and auditory perception. Present work aims to study the functional changes in cortical regions; through frequency domain features of EEG, in correlation to perception and attention level (visual and auditory), using a computer screen-control task. Psychophysiological data consisting nine channel EEG for 26 subjects were recorded during three conditions; 5 minutes eyes closed (EC), 5 minutes visual focus on a red dot on dark screen (DOT), and approximately 10 minutes battery test consisting 3 visual and 3 auditory subsections (intrinsic, cross-modal phasic and unimodal phasic) through perception and attention functions: alertness (WAFA). Relative power spectrum density (PSD) of different activity bands for each epoch of one second have been computed; corresponding to each task. For WAFA task, mean reaction time (MRT, in milliseconds) and dispersion of reaction time (DRT, in milliseconds) have been registered. Correlation values between PSD of every EC, DOT and WAFA; and, MRT and DRT of WAFA assignment was registered. For task EC, correlation values are positive for theta, alpha, negative and positive for MRT and DRT consecutively, throughout the cortical region. For task DOT, correlation values negative for delta, theta and gamma; positive for alpha and beta. For task WAFA, correlation values are negative for delta and theta, but for different subsections of visual and auditory (intrinsic, cross-modal phasic and unimodal phasic) are either of positive, negative or uncorrelated for alpha, beta and gamma throughout the cortical region. EEG band activities are observed to be directionally correlated with visual and auditory alertness level.
An EEG based Quantitative Analysis of Absorbed Meditative State G. Gaurav, Ashish Kumar Sahani, Abhijit Sahoo International IEEE EMBS Conference on Neural Engineering Ner, 2019 Meditation is a mental practice to achieve focus of mind and emotional clarity. Meditation has been used for cognitive enhancement, rehabilitation and reducing stress and anxiety. In the present study, we are doing a comparative analysis between various levels of meditators based on EEG as psychophysiological indicator; and possibility of EEG as a neurofeedback for meditators. An analytical experiment on three categories of subjects (A: an expert meditator, B: five moderate meditators and C: five non-meditators) was done. Each subject was guided to perform two visual tasks; first to sit relaxed with eyes closed (REC) and second to gaze on a dot on screen (RDOT); supplied, EEG being recorded in parallel. The first subject was recorded with absorbed state of meditation (Samādhi). For psychophysiological analysis, wavelet transform based features from each recording of EEG was evaluated. Topographical mapping of brain functioning based on features were plotted and analyzed. It was observed that theta, alpha and beta were comparatively higher for expert meditator in frontal and central region during REC and RDOT. Also, during absorbed meditative state, the alpha and beta are higher at midline central region (Cz) and theta is higher at C3 and C4.
Non-invasive EEG-metric based stress detection Gaurav, R. S. Anand, Vinod Kumar 4th IEEE International Conference on Signal Processing Computing and Control Ispcc 2017, 2017 Psychological stress is a vital parameter related to individual's health and cognitive performance which may affect emotions and professional efficiency. Regula stress profile generated can be used as neurofeedback for the clinical as personal assessment. This paper describes a method to detect mental stress level based on physiological parameters. In this method an electroencephalogram (EEG) parameter based binary stress classifier is developed which is validated through probabilistic stress profiler of differential stress inventory questionnaire. A non-invasive 9 channel EEG is used to extract physiological signal and an EEG-metric based cognitive state and workload outputs is generated for 41 healthy volunteers (37 males and 4 females, age; 24±5 years). All subjects were performed three simple tasks of closed eye, focusing vision on a red dot on center of dark screen and focusing on a white screen. Central tendencies (mean, median and mode) are extracted from of EEG-metric (sleep onset, distraction, low engagement, high engagement and cognitive states) as features. Either of the two classes as low stress or high stress are evaluated from probabilistic stress profiler of differential stress inventory and used as training output classes. A supervisory training of multiple layer perceptron based binary support vector machine classifier was used to detect stress class one by one. 40 subject's samples were used for training and interchanging one-by one 41th subject's stress class is determined from the designed classifier. Out of 41 subjects, stress level of 30 subjects is correctly identified.
Apnea sensing using photoplethysmography G Gaurav, S Mohanasankar, V Jagadeesh Kumar Proceedings of the International Conference on Sensing Technology Icst, 2013 Apnea (involuntary stoppage of breathing for a while) is a very common phenomenon. While in adults apnea may just disturb the sleep, for neonates apnea can be life threatening. This paper describes a technique of detecting apnea using photo-plethysmo-graphy (PPG). The proposed technique is easily applicable for monitoring neonates on a long term basis for apnea detection since PPG is a non invasive technique that can be used for long term monitoring of patients, including neonates. In the proposed method, the respiratory signal that is inscribed in a typical PPG is extracted using wavelet decomposition. The onset of apnea is then detected by applying an algorithm on the power spectrum of the extracted respiratory signal. The apnea detection algorithm proffered was tested on 16 subjects and the maximum time taken for detection of onset of apnea is found to be less than 10 s (except for an outlier of 15 s), which is clinically acceptable.
RECENT SCHOLAR PUBLICATIONS
A Machine Learning Approach to the Smartwatch-based Epileptic Seizure Detection System G Gaurav, R Shukla, G Singh, AK Sahani IETE Journal of Research , 2022 2022 Citations: 17
Epileptic seizure detection using continuous wavelet transform and deep neural networks R Shukla, B Kumar, G Gaurav, G Singh, AK Sahani Sensing Technology: Proceedings of ICST 2022, 291-300 , 2022 2022 Citations: 12
A frequency analysis-based apnea detection algorithm using photoplethysmography G Gaurav, RS Anand Computational Intelligence in Healthcare Applications, 197-208 , 2022 2022 Citations: 1
EEG based cognitive task classification using multifractal detrended fluctuation analysis G Gaurav, RS Anand, V Kumar Cognitive Neurodynamics 15 (6), 999-1013 , 2021 2021 Citations: 49
Epileptic seizure detection using STFT based peak mean feature and support vector machine N Sharma, G Gaurav, RS Anand 2021 8th International Conference on Signal Processing and Integrated … , 2021 2021 Citations: 4
Correlating visual and auditory perception and attention level using EEG parameters G Gaurav, RS Anand, V Kumar 2020 7th International Conference on Signal Processing and Integrated … , 2020 2020 Citations: 1
An eeg based quantitative analysis of absorbed meditative state G Gaurav, AK Sahani, A Sahoo 2019 9th International IEEE/EMBS Conference on Neural Engineering (NER), 57-60 , 2019 2019 Citations: 5
EEG-metric based mental stress detection Gaurav, RS Anand, V Kumar Network Biol , 2018 2018 Citations: 24
Non-invasive EEG-metric based stress detection Gaurav, RS Anand, V Kumar 2017 4th International Conference on Signal Processing, Computing and … , 2017 2017 Citations: 7
Apnea sensing using photoplethysmography G Gaurav, S Mohanasankar, VJ Kumar 2013 Seventh International Conference on Sensing Technology (ICST), 285-288 , 2013 2013 Citations: 12
MOST CITED SCHOLAR PUBLICATIONS
EEG based cognitive task classification using multifractal detrended fluctuation analysis G Gaurav, RS Anand, V Kumar Cognitive Neurodynamics 15 (6), 999-1013 , 2021 2021 Citations: 49
EEG-metric based mental stress detection Gaurav, RS Anand, V Kumar Network Biol , 2018 2018 Citations: 24
A Machine Learning Approach to the Smartwatch-based Epileptic Seizure Detection System G Gaurav, R Shukla, G Singh, AK Sahani IETE Journal of Research , 2022 2022 Citations: 17
Epileptic seizure detection using continuous wavelet transform and deep neural networks R Shukla, B Kumar, G Gaurav, G Singh, AK Sahani Sensing Technology: Proceedings of ICST 2022, 291-300 , 2022 2022 Citations: 12
Apnea sensing using photoplethysmography G Gaurav, S Mohanasankar, VJ Kumar 2013 Seventh International Conference on Sensing Technology (ICST), 285-288 , 2013 2013 Citations: 12
Non-invasive EEG-metric based stress detection Gaurav, RS Anand, V Kumar 2017 4th International Conference on Signal Processing, Computing and … , 2017 2017 Citations: 7
An eeg based quantitative analysis of absorbed meditative state G Gaurav, AK Sahani, A Sahoo 2019 9th International IEEE/EMBS Conference on Neural Engineering (NER), 57-60 , 2019 2019 Citations: 5
Epileptic seizure detection using STFT based peak mean feature and support vector machine N Sharma, G Gaurav, RS Anand 2021 8th International Conference on Signal Processing and Integrated … , 2021 2021 Citations: 4
A frequency analysis-based apnea detection algorithm using photoplethysmography G Gaurav, RS Anand Computational Intelligence in Healthcare Applications, 197-208 , 2022 2022 Citations: 1
Correlating visual and auditory perception and attention level using EEG parameters G Gaurav, RS Anand, V Kumar 2020 7th International Conference on Signal Processing and Integrated … , 2020 2020 Citations: 1
