Luis Alfredo Moctezuma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Luis Alfredo Moctezuma has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
Marta Molinas
Norwegian University of Science and Technology
7 shared publications- Marta Molinas
Department of Engineering Cybernetics, Norwegian University of Science and Technology, Norway
4 shared publications Takashi Abe
International Institute for Integrative Sleep Medicine (WPI-IIIS), University of Tsukuba, Tsukuba, Japan
3 shared publications- Takashi Abe
International Institute for Integrative Sleep Medicine (WPI-IIIS), University of Tsukuba, Tsukuba, Ibaraki, Japan
3 shared publications - Andres Soler
Norwegian University of Science and Technology
1 shared publication - Eduardo Giraldo
Research Group in Automatic Control, Electrical Engineering Department, Universidad Tecnológica de Pereira, Pereira 660003, Colombia
1 shared publication
Latest Publications
- 2 July 2026
- Published by Springer Nature
- p. 99-109
Emotion Recognition Across Wakefulness and Sleep Using High-Density EEG and Machine Learning Techniques
- 2 July 2026
- Published by Springer Nature
- p. 99-109
Emotion Recognition Across Wakefulness and Sleep Using High-Density EEG and Machine Learning Techniques
- Open Access
- 29 September 2025
- Published by Oxford University Press (OUP) in SLEEP Advances
- Vol. 6 (4) , zpaf066
From high- to low-density EEG for automatic classification of dream experiences during stage 2 of NREM
This study proposes a method to automatically identify dream experience (DE) and no experience (NE) during the sleep stage N2 of nonrapid eye movement (NREM). We investigated the use of machine learning (ML) to automatically identify when a subject is having a dream during NREM from electroencephalography (EEG) signals. We use permutation-based channel selection to identify the most informative EEG channels for the classification of DE and NE and to select a set of channels that allow focus on the most important areas of the brain at the scalp level. The results show that when the ML models are trained on a balanced dataset containing both DE and NE reports, along with high-density EEG, they can achieve a classification performance of up to 0.94 in accuracy, F1 score, precision and recall, an Area Under the Receiver Operating Characteristic of 0.97, and a kappa of 0.88. Performance decreases while we reduce the number of channels, but it remains like using up to 30–40 EEG channels. We show that ML models trained on high-density EEG to classify NE and DE can identify whether a subject was dreaming, achieving an accuracy of 0.7 on a separate set of dream reports where subjects reported a dream experience without recall, and channel selection methods have shown that performance could increase by 0.02 when EEG channels are removed from the occipital area. Our results show a high classification performance for automatic dream detection and the need to reduce the number of EEG channels needed to create the ML models, thus obtaining low-cost portable devices that can be used in real-life scenarios. Statement of Significance Dreaming during non-REM sleep is difficult to detect objectively. Here, we show that machine learning models leveraging high-density EEG can accurately classify the presence vs absence of dream experiences during N2 sleep. A key finding is that this high performance is maintained with a substantially reduced channel set, establishing the feasibility of portable, low-density EEG systems for real-time dream detection. Our work provides a critical objective correlate for subjective dreaming and opens new avenues for practical applications in sleep research and clinical diagnostics.
- 29 September 2025
- Published by Oxford University Press (OUP) in SLEEP Advances
- Vol. 6 (4) , zpaf066
Open AccessFrom high- to low-density EEG for automatic classification of dream experiences during stage 2 of NREM
This study proposes a method to automatically identify dream experience (DE) and no experience (NE) during the sleep stage N2 of nonrapid eye movement (NREM). We investigated the use of machine learning (ML) to automatically identify when a subject is having a dream during NREM from electroencephalography (EEG) signals. We use permutation-based channel selection to identify the most informative EEG channels for the classification of DE and NE and to select a set of channels that allow focus on the most important areas of the brain at the scalp level. The results show that when the ML models are trained on a balanced dataset containing both DE and NE reports, along with high-density EEG, they can achieve a classification performance of up to 0.94 in accuracy, F1 score, precision and recall, an Area Under the Receiver Operating Characteristic of 0.97, and a kappa of 0.88. Performance decreases while we reduce the number of channels, but it remains like using up to 30–40 EEG channels. We show that ML models trained on high-density EEG to classify NE and DE can identify whether a subject was dreaming, achieving an accuracy of 0.7 on a separate set of dream reports where subjects reported a dream experience without recall, and channel selection methods have shown that performance could increase by 0.02 when EEG channels are removed from the occipital area. Our results show a high classification performance for automatic dream detection and the need to reduce the number of EEG channels needed to create the ML models, thus obtaining low-cost portable devices that can be used in real-life scenarios. Statement of Significance Dreaming during non-REM sleep is difficult to detect objectively. Here, we show that machine learning models leveraging high-density EEG can accurately classify the presence vs absence of dream experiences during N2 sleep. A key finding is that this high performance is maintained with a substantially reduced channel set, establishing the feasibility of portable, low-density EEG systems for real-time dream detection. Our work provides a critical objective correlate for subjective dreaming and opens new avenues for practical applications in sleep research and clinical diagnostics.
- Open Access
- 1 January 2025
- Published by Wiley in BioMed Research International
- Vol. 2025 (1) , 3585125
Unlocking Dreams and Dreamless Sleep: Machine Learning Classification With Optimal EEG Channels
Research suggests that dreams play a role in the regulation of emotional processing and memory consolidation; electroencephalography (EEG) is useful for studying them, but manual annotation is time‐consuming and prone to bias. This study was aimed at developing an EEG‐based machine learning (ML) model to automatically identify dream and dreamless states in sleep. We extracted features from EEG data using common spatial patterns (CSPs) and the discrete wavelet transform (DWT) and used them to classify EEG signals into dream and dreamless states using ML models. To determine the most informative channels for classification, we used the permutation‐based channel selection method and the nondominated sorting genetic algorithm II (NSGA‐II). We evaluated our proposal using a public dataset that is part of the DREAM project, which was collected from 58 EEG channels during rapid eye movement (REM) and non‐REM sleep, while 28 subjects reported dream or dreamless experiences. We achieved accuracies greater than 0.85 to distinguish dream and dreamless states using CSP‐based feature extraction combined with k ‐nearest neighbors (KNN), as well as through multiple combinations of EEG channels identified by channel selection methods. Our findings suggest that as few as 8–10 EEG channels may be sufficient for dream recognition. Excluding one subject at a time during model training revealed challenges in generalizing the models to unseen subjects. Channel selection methods have proven to be effective in selecting relevant subsets of EEG channels to classify dreams and dreamless experiences. Our results demonstrate the feasibility of automatic dream detection and highlight the need to improve ML generalization.
- 1 January 2025
- Published by Wiley in BioMed Research International
- Vol. 2025 (1) , 3585125
Open AccessUnlocking Dreams and Dreamless Sleep: Machine Learning Classification With Optimal EEG Channels
Research suggests that dreams play a role in the regulation of emotional processing and memory consolidation; electroencephalography (EEG) is useful for studying them, but manual annotation is time‐consuming and prone to bias. This study was aimed at developing an EEG‐based machine learning (ML) model to automatically identify dream and dreamless states in sleep. We extracted features from EEG data using common spatial patterns (CSPs) and the discrete wavelet transform (DWT) and used them to classify EEG signals into dream and dreamless states using ML models. To determine the most informative channels for classification, we used the permutation‐based channel selection method and the nondominated sorting genetic algorithm II (NSGA‐II). We evaluated our proposal using a public dataset that is part of the DREAM project, which was collected from 58 EEG channels during rapid eye movement (REM) and non‐REM sleep, while 28 subjects reported dream or dreamless experiences. We achieved accuracies greater than 0.85 to distinguish dream and dreamless states using CSP‐based feature extraction combined with k ‐nearest neighbors (KNN), as well as through multiple combinations of EEG channels identified by channel selection methods. Our findings suggest that as few as 8–10 EEG channels may be sufficient for dream recognition. Excluding one subject at a time during model training revealed challenges in generalizing the models to unseen subjects. Channel selection methods have proven to be effective in selecting relevant subsets of EEG channels to classify dreams and dreamless experiences. Our results demonstrate the feasibility of automatic dream detection and highlight the need to improve ML generalization.
- Open Access
- 2 August 2024
- Published by Springer Nature in Scientific Reports
- Vol. 14 (1) , 1-16
GRU-powered sleep stage classification with permutation-based EEG channel selection
We present a new approach to classifying the sleep stage that incorporates a computationally inexpensive method based on permutations for channel selection and takes advantage of deep learning power, specifically the gated recurrent unit (GRU) model, along with other deep learning methods. By systematically permuting the electroencephalographic (EEG) channels, different combinations of EEG channels are evaluated to identify the most informative subset for the classification of the 5-class sleep stage. For analysis, we used an EEG dataset that was collected at the International Institute for Integrative Sleep Medicine (WPI-IIIS) at the University of Tsukuba in Japan. The results of these explorations provide many new insights such as the (1) drastic decrease in performance when channels are fewer than 3, (2) 3-random channels selected by permutation provide the same or better prediction than the 3 channels recommended by the American Academy of Sleep Medicine (AASM), (3) N1 class suffers the most in prediction accuracy as the channels drop from 128 to 3 random or 3 AASM, and (4) no single channel provides acceptable levels of accuracy in the prediction of 5 classes. The results obtained show the GRU’s ability to retain essential temporal information from EEG data, which allows capturing the underlying patterns associated with each sleep stage effectively. Using permutation-based channel selection, we enhance or at least maintain as high model efficiency as when using high-density EEG, incorporating only the most informative EEG channels.
- 2 August 2024
- Published by Springer Nature in Scientific Reports
- Vol. 14 (1) , 1-16
Open AccessGRU-powered sleep stage classification with permutation-based EEG channel selection
We present a new approach to classifying the sleep stage that incorporates a computationally inexpensive method based on permutations for channel selection and takes advantage of deep learning power, specifically the gated recurrent unit (GRU) model, along with other deep learning methods. By systematically permuting the electroencephalographic (EEG) channels, different combinations of EEG channels are evaluated to identify the most informative subset for the classification of the 5-class sleep stage. For analysis, we used an EEG dataset that was collected at the International Institute for Integrative Sleep Medicine (WPI-IIIS) at the University of Tsukuba in Japan. The results of these explorations provide many new insights such as the (1) drastic decrease in performance when channels are fewer than 3, (2) 3-random channels selected by permutation provide the same or better prediction than the 3 channels recommended by the American Academy of Sleep Medicine (AASM), (3) N1 class suffers the most in prediction accuracy as the channels drop from 128 to 3 random or 3 AASM, and (4) no single channel provides acceptable levels of accuracy in the prediction of 5 classes. The results obtained show the GRU’s ability to retain essential temporal information from EEG data, which allows capturing the underlying patterns associated with each sleep stage effectively. Using permutation-based channel selection, we enhance or at least maintain as high model efficiency as when using high-density EEG, incorporating only the most informative EEG channels.
- 13 September 2023
- Published by Springer Nature
- p. 71-82
Decoding Emotion Dimensions Arousal and Valence Elicited on EEG Responses to Videos and Images: A Comparative Evaluation
- 13 September 2023
- Published by Springer Nature
- p. 71-82
Decoding Emotion Dimensions Arousal and Valence Elicited on EEG Responses to Videos and Images: A Comparative Evaluation
- Open Access
- 3 March 2022
- Published by Springer Nature in Scientific Reports
- Vol. 12 (1) , 1-15
Two-dimensional CNN-based distinction of human emotions from EEG channels selected by multi-objective evolutionary algorithm
In this study we explore how different levels of emotional intensity (Arousal) and pleasantness (Valence) are reflected in electroencephalographic (EEG) signals. We performed the experiments on EEG data of 32 subjects from the DEAP public dataset, where the subjects were stimulated using 60-s videos to elicitate different levels of Arousal/Valence and then self-reported the rating from 1 to 9 using the self-assessment Manikin (SAM). The EEG data was pre-processed and used as input to a convolutional neural network (CNN). First, the 32 EEG channels were used to compute the maximum accuracy level obtainable for each subject as well as for creating a single model using data from all the subjects. The experiment was repeated using one channel at a time, to see if specific channels contain more information to discriminate between low vs high arousal/valence. The results indicate than using one channel the accuracy is lower compared to using all the 32 channels. An optimization process for EEG channel selection is then designed with the Non-dominated Sorting Genetic Algorithm II (NSGA-II) with the objective to obtain optimal channel combinations with high accuracy recognition. The genetic algorithm evaluates all possible combinations using a chromosome representation for all the 32 channels, and the EEG data from each chromosome in the different populations are tested iteratively solving two unconstrained objectives; to maximize classification accuracy and to reduce the number of required EEG channels for the classification process. Best combinations obtained from a Pareto-front suggests that as few as 8–10 channels can fulfill this condition and provide the basis for a lighter design of EEG systems for emotion recognition. In the best case, the results show accuracies of up to 1.00 forlow vs high arousalusing eight EEG channels, and 1.00 forlow vs high valenceusing only two EEG channels. These results are encouraging for research and healthcare applications that will require automatic emotion recognition with wearable EEG.
- 3 March 2022
- Published by Springer Nature in Scientific Reports
- Vol. 12 (1) , 1-15
Open AccessTwo-dimensional CNN-based distinction of human emotions from EEG channels selected by multi-objective evolutionary algorithm
In this study we explore how different levels of emotional intensity (Arousal) and pleasantness (Valence) are reflected in electroencephalographic (EEG) signals. We performed the experiments on EEG data of 32 subjects from the DEAP public dataset, where the subjects were stimulated using 60-s videos to elicitate different levels of Arousal/Valence and then self-reported the rating from 1 to 9 using the self-assessment Manikin (SAM). The EEG data was pre-processed and used as input to a convolutional neural network (CNN). First, the 32 EEG channels were used to compute the maximum accuracy level obtainable for each subject as well as for creating a single model using data from all the subjects. The experiment was repeated using one channel at a time, to see if specific channels contain more information to discriminate between low vs high arousal/valence. The results indicate than using one channel the accuracy is lower compared to using all the 32 channels. An optimization process for EEG channel selection is then designed with the Non-dominated Sorting Genetic Algorithm II (NSGA-II) with the objective to obtain optimal channel combinations with high accuracy recognition. The genetic algorithm evaluates all possible combinations using a chromosome representation for all the 32 channels, and the EEG data from each chromosome in the different populations are tested iteratively solving two unconstrained objectives; to maximize classification accuracy and to reduce the number of required EEG channels for the classification process. Best combinations obtained from a Pareto-front suggests that as few as 8–10 channels can fulfill this condition and provide the basis for a lighter design of EEG systems for emotion recognition. In the best case, the results show accuracies of up to 1.00 forlow vs high arousalusing eight EEG channels, and 1.00 forlow vs high valenceusing only two EEG channels. These results are encouraging for research and healthcare applications that will require automatic emotion recognition with wearable EEG.
- Open Access
Automated methodology for optimal selection of minimum electrode subsets for accurate EEG source estimation based on Genetic Algorithm optimization
High-density Electroencephalography (HD-EEG) has been proven to be the most accurate option to estimate the neural activity inside the brain. Multiple studies report the effect of electrode number on source localization for specific sources and specific electrode configurations. The electrodes for these configurations have been manually selected to uniformly cover the entire head, going from 32 to 128 electrodes, but electrodes were not selected according to their contribution to accuracy. In this work, an optimization-based study is proposed to determine the minimum number of electrodes and identify optimal combinations of electrodes that can retain the localization accuracy of HD-EEG reconstructions. This optimization approach incorporates scalp landmark positions of widely used EEG montages. In this way, a systematic search for the minimum electrode subset is performed for single and multiple source localization problems. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) combined with source reconstruction methods is used to formulate a multi-objective optimization problem that concurrently minimizes 1) the localization error for each source and 2) the number of required EEG electrodes. The method can be used for evaluating the source localization quality of low-density EEG systems (e.g. consumer-grade wearable EEG). We performed an evaluation over synthetic and real EEG dataset with known ground-truth. The experimental results show that optimal subsets with 6 electrodes can obtain an equal or better accuracy than HD-EEG (with more than 200 channels) for a single source case. This happened when reconstructing a particular brain activity in more than 88% of the cases (in synthetic signals) and 63% (in real signals), and in more than 88% and 73% of the cases when considering optimal combinations with 8 channels. For a multiple-source case of three sources (only with synthetic signals), it was found that optimized combinations of 8, 12 and 16 electrodes attained an equal or better accuracy than HD-EEG with 231 electrodes in at least 58%, 76%, and 82% of the cases respectively. Additionally, for such electrode numbers, a lower mean error and standard deviation than with 231 electrodes were obtained. Highlights: The number of EEG electrodes and their locations can be optimized for reconstructing the brain source activity. Optimally selected EEG electrodes can retain the accuracy of high density montages (256, 128 chs) for brain source estimation, when electrodes are selected according to their contribution to accuracy. With optimization, selected combinations of EEG electrodes will flexibilize the estimation of the source activity.
Open AccessAutomated methodology for optimal selection of minimum electrode subsets for accurate EEG source estimation based on Genetic Algorithm optimization
High-density Electroencephalography (HD-EEG) has been proven to be the most accurate option to estimate the neural activity inside the brain. Multiple studies report the effect of electrode number on source localization for specific sources and specific electrode configurations. The electrodes for these configurations have been manually selected to uniformly cover the entire head, going from 32 to 128 electrodes, but electrodes were not selected according to their contribution to accuracy. In this work, an optimization-based study is proposed to determine the minimum number of electrodes and identify optimal combinations of electrodes that can retain the localization accuracy of HD-EEG reconstructions. This optimization approach incorporates scalp landmark positions of widely used EEG montages. In this way, a systematic search for the minimum electrode subset is performed for single and multiple source localization problems. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) combined with source reconstruction methods is used to formulate a multi-objective optimization problem that concurrently minimizes 1) the localization error for each source and 2) the number of required EEG electrodes. The method can be used for evaluating the source localization quality of low-density EEG systems (e.g. consumer-grade wearable EEG). We performed an evaluation over synthetic and real EEG dataset with known ground-truth. The experimental results show that optimal subsets with 6 electrodes can obtain an equal or better accuracy than HD-EEG (with more than 200 channels) for a single source case. This happened when reconstructing a particular brain activity in more than 88% of the cases (in synthetic signals) and 63% (in real signals), and in more than 88% and 73% of the cases when considering optimal combinations with 8 channels. For a multiple-source case of three sources (only with synthetic signals), it was found that optimized combinations of 8, 12 and 16 electrodes attained an equal or better accuracy than HD-EEG with 231 electrodes in at least 58%, 76%, and 82% of the cases respectively. Additionally, for such electrode numbers, a lower mean error and standard deviation than with 231 electrodes were obtained. Highlights: The number of EEG electrodes and their locations can be optimized for reconstructing the brain source activity. Optimally selected EEG electrodes can retain the accuracy of high density montages (256, 128 chs) for brain source estimation, when electrodes are selected according to their contribution to accuracy. With optimization, selected combinations of EEG electrodes will flexibilize the estimation of the source activity.
- Open Access
- 1 January 2020
- Published by Journal of Biomedical Research in Journal of Biomedical Research
- Vol. 34 (3) , 180-190
Classification of low-density EEG for epileptic seizures by energy and fractal features based on EMD
We are here to present a new method for the classification of epileptic seizures from electroencephalogram (EEG) signals. It consists of applying empirical mode decomposition (EMD) to extract the most relevant intrinsic mode functions (IMFs) and subsequent computation of the Teager and instantaneous energy, Higuchi and Petrosian fractal dimension, and detrended fluctuation analysis (DFA) for each IMF. We validated the method using a public dataset of 24 subjects with EEG signals from 22 channels and showed that it is possible to classify the epileptic seizures, even with segments of six seconds and a smaller number of channels (e.g., an accuracy of 0.93 using five channels). We were able to create a general machine-learning-based model to detect epileptic seizures of new subjects using epileptic-seizure data from various subjects, after reducing the number of instances, based on the k-means algorithm.
- 1 January 2020
- Published by Journal of Biomedical Research in Journal of Biomedical Research
- Vol. 34 (3) , 180-190
Open AccessClassification of low-density EEG for epileptic seizures by energy and fractal features based on EMD
We are here to present a new method for the classification of epileptic seizures from electroencephalogram (EEG) signals. It consists of applying empirical mode decomposition (EMD) to extract the most relevant intrinsic mode functions (IMFs) and subsequent computation of the Teager and instantaneous energy, Higuchi and Petrosian fractal dimension, and detrended fluctuation analysis (DFA) for each IMF. We validated the method using a public dataset of 24 subjects with EEG signals from 22 channels and showed that it is possible to classify the epileptic seizures, even with segments of six seconds and a smaller number of channels (e.g., an accuracy of 0.93 using five channels). We were able to create a general machine-learning-based model to detect epileptic seizures of new subjects using epileptic-seizure data from various subjects, after reducing the number of instances, based on the k-means algorithm.
- 2 February 2019
- Published by Springer Nature
- p. 830-846
Subject Identification from Low-Density EEG-Recordings of Resting-States: A Study of Feature Extraction and Classification
- 2 February 2019
- Published by Springer Nature
- p. 830-846
Subject Identification from Low-Density EEG-Recordings of Resting-States: A Study of Feature Extraction and Classification
- Open Access
- 1 February 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Sex differences observed in a study of EEG of linguistic activity and resting-state: Exploring optimal EEG channel configurations
This study reports the differences observed in the EEG signals of linguistic activity and resting-state between male and female subjects in a population of 16 individuals (8 females and 8 males). These differences were spotted while performing two experiments: sex identification and subject identification, where the initial aim was to identify the optimal number and placement of EEG channels to obtain high accuracies in sex and subject identification. The results of the identification show that the signals analyzed contain sex-specific information and that the best features from this sex-specific information are extracted from different EEG channel locations and from different hemispheres of the brain, for either sex. The effect of the number of electrodes and electrode localization is seen with clear differences between male and female subjects. The accuracy loss for sex identification when reducing the number of channels from 14 to 1 was of only 0.03 points during resting states (Accuracies from 0.79 to 0.76). For subject identification within either male or female groups during resting states, the accuracy loss was larger when reducing the number of channels from 14 to 1 (0.96 to 0.71 for female, 0.96 to 0.81 for male subjects). One finding of this study is that Theta and Gamma bands are strongest for males in the right hemisphere during resting states, whereas during linguistic activity these bands exhibit similar strengths in the left hemisphere for both males and females. Similar specific features in brain signals may enable the design of a flexible EEG device that can be adapted to specific mental tasks and Subject settings.
- 1 February 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Open AccessSex differences observed in a study of EEG of linguistic activity and resting-state: Exploring optimal EEG channel configurations
This study reports the differences observed in the EEG signals of linguistic activity and resting-state between male and female subjects in a population of 16 individuals (8 females and 8 males). These differences were spotted while performing two experiments: sex identification and subject identification, where the initial aim was to identify the optimal number and placement of EEG channels to obtain high accuracies in sex and subject identification. The results of the identification show that the signals analyzed contain sex-specific information and that the best features from this sex-specific information are extracted from different EEG channel locations and from different hemispheres of the brain, for either sex. The effect of the number of electrodes and electrode localization is seen with clear differences between male and female subjects. The accuracy loss for sex identification when reducing the number of channels from 14 to 1 was of only 0.03 points during resting states (Accuracies from 0.79 to 0.76). For subject identification within either male or female groups during resting states, the accuracy loss was larger when reducing the number of channels from 14 to 1 (0.96 to 0.71 for female, 0.96 to 0.81 for male subjects). One finding of this study is that Theta and Gamma bands are strongest for males in the right hemisphere during resting states, whereas during linguistic activity these bands exhibit similar strengths in the left hemisphere for both males and females. Similar specific features in brain signals may enable the design of a flexible EEG device that can be adapted to specific mental tasks and Subject settings.