Knowledge and Perception of the Use of AI and its Implementation in the Field of Radiology: Cross-Sectional Study.

BACKGROUND: Artificial Intelligence (AI) has been developing for decades, but in recent years its use in the field of health care has experienced an exponential increase. Currently, there is little doubt that these tools have transformed clinical practice. Therefore, it is important to know how the population perceives its implementation to be able to propose strategies for acceptance and implementation and to improve or prevent problems arising from future applications. OBJECTIVE: This study aims to describe the population's perception and knowledge of the use of AI as a health support tool and its application to radiology through a validated questionnaire, in order to develop strategies aimed at increasing acceptance of AI use, reducing possible resistance to change and identifying possible sociodemographic factors related to perception and knowledge. METHODS: A cross-sectional observational study was conducted using an anonymous and voluntarily validated questionnaire aimed at the entire population of Catalonia aged 18 years or older. The survey addresses 4 dimensions defined to describe users' perception of the use of AI in radiology, (1) "distrust and accountability," (2) "personal interaction," (3) "efficiency," and (4) "being informed," all with questions in a Likert scale format. Results closer to 5 refer to a negative perception of the use of AI, while results closer to 1 express a positive perception. Univariate and bivariate analyses were performed to assess possible associations between the 4 dimensions and sociodemographic characteristics. RESULTS: A total of 379 users responded to the survey, with an average age of 43.9 (SD 17.52) years and 59.8% (n=226) of them identified as female. In addition, 89.8% (n=335) of respondents indicated that they understood the concept of AI. Of the 4 dimensions analyzed, "distrust and accountability" obtained a mean score of 3.37 (SD 0.53), "personal interaction" obtained a mean score of 4.37 (SD 0.60), "efficiency" obtained a mean score of 3.06 (SD 0.73) and "being informed" obtained a mean score of 3.67 (SD 0.57). In relation to the "distrust and accountability" dimension, women, people older than 65 years, the group with university studies, and the population that indicated not understanding the AI concept had significantly more distrust in the use of AI. On the dimension of "being informed," it was observed that the group with university studies rated access to information more positively and those who indicated not understanding the concept of AI rated it more negatively. CONCLUSIONS: The majority of the sample investigated reported being familiar with the concept of AI, with varying degrees of acceptance of its implementation in radiology. It is clear that the most conflictive dimension is "personal interaction," whereas "efficiency" is where there is the greatest acceptance, being the dimension in which there are the best expectations for the implementation of AI in radiology.

Special Issue "Machine Learning Methods for Biomedical Data Analysis".

Machine learning is an effective method for developing automatic algorithms for analysing sophisticated biomedical data [...].

Effect of Foot Orthoses on Angular Velocity of Feet.

There is some uncertainty regarding how foot orthoses (FO) affect the biomechanics of the lower extremities during running in non-injured individuals. This study aims to describe the behavior of the angular velocity of the foot in the stride cycle measured with a low-sampling-rate IMU device commonly used by podiatrists. Specific objectives were to determine if there are differences in angular velocity between the right and left foot and to determine the effect of foot orthoses (FO) on the 3D angular velocity of the foot during running. The sample was composed of 40 male adults (age: 43.0 ± 13.8 years, weight: 72.0 ± 5.5 kg, and height: 175.5 ± 7.0 cm), who were healthy and without any locomotor system alterations at the time of the test. All subjects use FO on a regular basis. The results show that there are significant differences in the transverse plane between feet, with greater differences in the right foot. Significant differences between FO and non-FO conditions were observed in the frontal and transverse planes on the left foot and in the sagittal and transverse planes on the right foot. FO decreases the velocity of the foot in dorsi-plantar flexion and abduction and increases the velocity in inversion. The kinematic changes in foot velocity occur between 30% and 60% of the complete cycle, and the FO reduces the velocity in abduction and dorsi-plantar flexion and increases the velocity in inversion-eversion, which facilitates the transition to the oscillating leg and with it the displacement of the center of mass. Quantifying possible asymmetries and assessing the effect of foot orthoses may aid in improving running mechanics and preventing injuries in individuals.

Cerebral cortex layer segmentation using diffusion magnetic resonance imaging in vivo with applications to laminar connections and working memory analysis.

Understanding the laminar brain structure is of great help in further developing our knowledge of the functions of the brain. However, since most layer segmentation methods are invasive, it is difficult to apply them to the human brain in vivo. To systematically explore the human brain's laminar structure noninvasively, the K-means clustering algorithm was used to automatically segment the left hemisphere into two layers, the superficial and deep layers, using a 7 Tesla (T) diffusion magnetic resonance imaging (dMRI)open dataset. The obtained layer thickness was then compared with the layer thickness of the BigBrain reference dataset, which segmented the neocortex into six layers based on the von Economo atlas. The results show a significant correlation not only between our automatically segmented superficial layer thickness and the thickness of layers 1-3 from the reference histological data, but also between our automatically segmented deep layer thickness and the thickness of layers 4-6 from the reference histological data. Second, we constructed the laminar connections between two pairs of unidirectional connected regions, which is consistent with prior research. Finally, we conducted the laminar analysis of the working memory, which was challenging to do in the past, and explained the conclusions of the functional analysis. Our work successfully demonstrates that it is possible to segment the human cortex noninvasively into layers using dMRI data and further explores the mechanisms of the human brain.

A Machine Learning Approach for Predicting Non-Suicidal Self-Injury in Young Adults.

Artificial intelligence techniques were explored to assess the ability to anticipate self-harming behaviour in the mental health context using a database collected by an app previously designed to record the emotional states and activities of a group of subjects exhibiting self-harm. Specifically, the Leave-One-Subject-Out technique was used to train classification trees with a maximum of five splits. The results show an accuracy of 84.78%, a sensitivity of 64.64% and a specificity of 85.53%. In addition, positive and negative predictive values were also obtained, with results of 14.48% and 98.47%, respectively. These results are in line with those reported in previous work using a multilevel mixed-effect regression analysis. The combination of apps and AI techniques is a powerful way to improve the tools to accompany and support the care and treatment of patients with this type of behaviour. These studies also guide the improvement of apps on the user side, simplifying and collecting more meaningful data, and on the therapist side, progressing in pathology treatments. Traditional therapy involves observing and reconstructing what had happened before episodes once they have occurred. This new generation of tools will make it possible to monitor the pathology more closely and to act preventively.

IMU-Based Effects Assessment of the Use of Foot Orthoses in the Stance Phase during Running and Asymmetry between Extremities.

The objectives of this study were to determine the amplitude of movement differences and asymmetries between feet during the stance phase and to evaluate the effects of foot orthoses (FOs) on foot kinematics in the stance phase during running. In total, 40 males were recruited (age: 43.0 ± 13.8 years, weight: 72.0 ± 5.5 kg, height: 175.5 ± 7.0 cm). Participants ran on a running treadmill at 2.5 m/s using their own footwear, with and without the FOs. Two inertial sensors fixed on the instep of each of the participant's footwear were used. Amplitude of movement along each axis, contact time and number of steps were considered in the analysis. The results indicate that the movement in the sagittal plane is symmetric, but that it is not in the frontal and transverse planes. The right foot displayed more degrees of movement amplitude than the left foot although these differences are only significant in the abduction case. When FOs are used, a decrease in amplitude of movement in the three axes is observed, except for the dorsi-plantar flexion in the left foot and both feet combined. The contact time and the total step time show a significant increase when FOs are used, but the number of steps is not altered, suggesting that FOs do not interfere in running technique. The reduction in the amplitude of movement would indicate that FOs could be used as a preventive tool. The FOs do not influence the asymmetry of the amplitude of movement observed between feet, and this risk factor is maintained. IMU devices are useful tools to detect risk factors related to running injuries. With its use, even more personalized FOs could be manufactured.

Clustering Users to Determine the Most Suitable Gamification Elements.

The use of gamification elements has extended from being a complement for a product to being integrated into multiple public services to motivate the user. The first drawback for service designers is choosing which gamification elements are appropriate for the intended audience, in addition to the possible incompatibilities between gamification elements. This work proposes a clustering technique that enables mapping different user profiles in relation to their preferred gamification elements. Additionally, by mapping the best cluster for each gamification element, it is possible to determine the preferred game genre. The article answered the following research questions: What is the relationship between the genre of the game and the element of gamification? Different user groups (profiles) for each gamification element? Results indicate that there are cases where the users are divided between those who agree or disagree. However, other elements present a great heterogeneity in the number of groups and the levels of agreement.

A Fast Approach to Removing Muscle Artifacts for EEG with Signal Serialization Based Ensemble Empirical Mode Decomposition.

An electroencephalogram (EEG) is an electrophysiological signal reflecting the functional state of the brain. As the control signal of the brain-computer interface (BCI), EEG may build a bridge between humans and computers to improve the life quality for patients with movement disorders. The collected EEG signals are extremely susceptible to the contamination of electromyography (EMG) artifacts, affecting their original characteristics. Therefore, EEG denoising is an essential preprocessing step in any BCI system. Previous studies have confirmed that the combination of ensemble empirical mode decomposition (EEMD) and canonical correlation analysis (CCA) can effectively suppress EMG artifacts. However, the time-consuming iterative process of EEMD may limit the application of the EEMD-CCA method in real-time monitoring of BCI. Compared with the existing EEMD, the recently proposed signal serialization based EEMD (sEEMD) is a good choice to provide effective signal analysis and fast mode decomposition. In this study, an EMG denoising method based on sEEMD and CCA is discussed. All of the analyses are carried out on semi-simulated data. The results show that, in terms of frequency and amplitude, the intrinsic mode functions (IMFs) decomposed by sEEMD are consistent with the IMFs obtained by EEMD. There is no significant difference in the ability to separate EMG artifacts from EEG signals between the sEEMD-CCA method and the EEMD-CCA method (p > 0.05). Even in the case of heavy contamination (signal-to-noise ratio is less than 2 dB), the relative root mean squared error is about 0.3, and the average correlation coefficient remains above 0.9. The running speed of the sEEMD-CCA method to remove EMG artifacts is significantly improved in comparison with that of EEMD-CCA method (p < 0.05). The running time of the sEEMD-CCA method for three lengths of semi-simulated data is shortened by more than 50%. This indicates that sEEMD-CCA is a promising tool for EMG artifact removal in real-time BCI systems.

Automatic Analysis of Archimedes' Spiral for Characterization of Genetic Essential Tremor Based on Shannon's Entropy and Fractal Dimension.

Among neural disorders related to movement, essential tremor has the highest prevalence; in fact, it is twenty times more common than Parkinson's disease. The drawing of the Archimedes' spiral is the gold standard test to distinguish between both pathologies. The aim of this paper is to select non-linear biomarkers based on the analysis of digital drawings. It belongs to a larger cross study for early diagnosis of essential tremor that also includes genetic information. The proposed automatic analysis system consists in a hybrid solution: Machine Learning paradigms and automatic selection of features based on statistical tests using medical criteria. Moreover, the selected biomarkers comprise not only commonly used linear features (static and dynamic), but also other non-linear ones: Shannon entropy and Fractal Dimension. The results are hopeful, and the developed tool can easily be adapted to users; and taking into account social and economic points of view, it could be very helpful in real complex environments.

Towards Semi-Automatic Artifact Rejection for the Improvement of Alzheimer's Disease Screening from EEG Signals.

A large number of studies have analyzed measurable changes that Alzheimer's disease causes on electroencephalography (EEG). Despite being easily reproducible, those markers have limited sensitivity, which reduces the interest of EEG as a screening tool for this pathology. This is for a large part due to the poor signal-to-noise ratio of EEG signals: EEG recordings are indeed usually corrupted by spurious extra-cerebral artifacts. These artifacts are responsible for a consequent degradation of the signal quality. We investigate the possibility to automatically clean a database of EEG recordings taken from patients suffering from Alzheimer's disease and healthy age-matched controls. We present here an investigation of commonly used markers of EEG artifacts: kurtosis, sample entropy, zero-crossing rate and fractal dimension. We investigate the reliability of the markers, by comparison with human labeling of sources. Our results show significant differences with the sample entropy marker. We present a strategy for semi-automatic cleaning based on blind source separation, which may improve the specificity of Alzheimer screening using EEG signals.

Wind turbine database for intelligent operation and maintenance strategies.

With the aim of helping researchers to develop intelligent operation and maintenance strategies, in this manuscript, an extensive 3-years Supervisory Control and Data Acquisition database of five Fuhrländer FL2500 2.5 MW wind turbines is presented. The database contains 312 analogous variables recorded at 5-minute intervals, from 78 different sensors. The reported values for each sensor are minimum, maximum, mean, and standard deviation. The database also contains the alarm events, indicating the system and subsystem and a small description. Finally, a set of functions to download specific subsets of the whole database is freely available in Matlab, R, and Python. To demonstrate the usefulness of this database, an illustrative example is given. In this example, different gearbox variables are selected to estimate a target variable to detect whether or not the estimate differs from the actual value provided for the sensor. By using this normality modelling approach, it is possible to detect rotor malfunction when the estimate differs from the actual measured value.

Early prediction of dementia using fMRI data with a graph convolutional network approach.

Objective. Alzheimer's disease is a progressive neurodegenerative dementia that poses a significant global health threat. It is imperative and essential to detect patients in the mild cognitive impairment (MCI) stage or even earlier, enabling effective interventions to prevent further deterioration of dementia. This study focuses on the early prediction of dementia utilizing Magnetic Resonance Imaging (MRI) data, using the proposed Graph Convolutional Networks (GCNs).Approach. Specifically, we developed a functional connectivity (FC) based GCN framework for binary classifications using resting-state fMRI data. We explored different types and processing methods of FC and evaluated the performance on the OASIS-3 dataset. We developed the GCN model for two different purposes: (1) MCI diagnosis: classifying MCI from normal controls (NCs); and (2) dementia risk prediction: classifying NCs from subjects who have the potential for developing MCI but have not been clinically diagnosed as MCI.Main results. The results of the experiments revealed several important findings: First, the proposed GCN outperformed both the baseline GCN and Support Vector Machine (SVM). It achieved the best average accuracy of 80.3% (11.7% higher than the baseline GCN and 23.5% higher than SVM) and the highest accuracy of 91.2%. Secondly, the GCN framework with (absolute) individual FC performed slightly better than that with global FC generally. However, GCN using global graphs with appropriate connectivity can achieve equivalent or superior performance to individual graphs in some cases, which highlights the significance of suitable connectivity for achieving performance. Additionally, the results indicate that the self-network connectivity of specific brain network regions (such as default mode network, visual network, ventral attention network and somatomotor network) may play a more significant role in GCN classification.Significance. Overall, this study offers valuable insights into the application of GCNs in brain analysis and early diagnosis of dementia. This contributes significantly to the understanding of MCI and has substantial potential for clinical applications in early diagnosis and intervention for dementia and other neurodegenerative diseases. Our code for GCN implementation is available at:https://github.com/Shuning-Han/FC-based-GCN.

Real-world testing of an artificial intelligence algorithm for the analysis of chest X-rays in primary care settings.

Interpreting chest X-rays is a complex task, and artificial intelligence algorithms for this purpose are currently being developed. It is important to perform external validations of these algorithms in order to implement them. This study therefore aims to externally validate an AI algorithm's diagnoses in real clinical practice, comparing them to a radiologist's diagnoses. The aim is also to identify diagnoses the algorithm may not have been trained for. A prospective observational study for the external validation of the AI algorithm in a region of Catalonia, comparing the AI algorithm's diagnosis with that of the reference radiologist, considered the gold standard. The external validation was performed with a sample of 278 images and reports, 51.8% of which showed no radiological abnormalities according to the radiologist's report. Analysing the validity of the AI algorithm, the average accuracy was 0.95 (95% CI 0.92; 0.98), the sensitivity was 0.48 (95% CI 0.30; 0.66) and the specificity was 0.98 (95% CI 0.97; 0.99). The conditions where the algorithm was most sensitive were external, upper abdominal and cardiac and/or valvular implants. On the other hand, the conditions where the algorithm was less sensitive were in the mediastinum, vessels and bone. The algorithm has been validated in the primary care setting and has proven to be useful when identifying images with or without conditions. However, in order to be a valuable tool to help and support experts, it requires additional real-world training to enhance its diagnostic capabilities for some of the conditions analysed. Our study emphasizes the need for continuous improvement to ensure the algorithm's effectiveness in primary care.

Real-time estimation of EEG-based engagement in different tasks.

Objective.Recent trends in brain-computer interface (BCI) research concern the passive monitoring of brain activity, which aim to monitor a wide variety of cognitive states. Engagement is such a cognitive state, which is of interest in contexts such as learning, entertainment or rehabilitation. This study proposes a novel approach for real-time estimation of engagement during different tasks using electroencephalography (EEG).Approach.Twenty-three healthy subjects participated in the BCI experiment. A modified version of the d2 test was used to elicit engagement. Within-subject classification models which discriminate between engaging and resting states were trained based on EEG recorded during a d2 test based paradigm. The EEG was recorded using eight electrodes and the classification model was based on filter-bank common spatial patterns and a linear discriminant analysis. The classification models were evaluated in cross-task applications, namely when playing Tetris at different speeds (i.e. slow, medium, fast) and when watching two videos (i.e. advertisement and landscape video). Additionally, subjects' perceived engagement was quantified using a questionnaire.Main results.The models achieved a classification accuracy of 90% on average when tested on an independent d2 test paradigm recording. Subjects' perceived and estimated engagement were found to be greater during the advertisement compared to the landscape video (p= 0.025 andp<0.001, respectively); greater during medium and fast compared to slow Tetris speed (p<0.001, respectively); not different between medium and fast Tetris speeds. Additionally, a common linear relationship was observed for perceived and estimated engagement (rrm= 0.44,p<0.001). Finally, theta and alpha band powers were investigated, which respectively increased and decreased during more engaging states.Significance.This study proposes a task-specific EEG engagement estimation model with cross-task capabilities, offering a framework for real-world applications.

On the selection of non-invasive methods based on speech analysis oriented to automatic Alzheimer disease diagnosis.

The work presented here is part of a larger study to identify novel technologies and biomarkers for early Alzheimer disease (AD) detection and it focuses on evaluating the suitability of a new approach for early AD diagnosis by non-invasive methods. The purpose is to examine in a pilot study the potential of applying intelligent algorithms to speech features obtained from suspected patients in order to contribute to the improvement of diagnosis of AD and its degree of severity. In this sense, Artificial Neural Networks (ANN) have been used for the automatic classification of the two classes (AD and control subjects). Two human issues have been analyzed for feature selection: Spontaneous Speech and Emotional Response. Not only linear features but also non-linear ones, such as Fractal Dimension, have been explored. The approach is non invasive, low cost and without any side effects. Obtained experimental results were very satisfactory and promising for early diagnosis and classification of AD patients.

Multi-Class Classification of Upper Limb Movements With Filter Bank Task-Related Component Analysis.

The classification of limb movements can provide with control commands in non-invasive brain-computer interface. Previous studies on the classification of limb movements have focused on the classification of left/right limbs; however, the classification of different types of upper limb movements has often been ignored despite that it provides more active-evoked control commands in the brain-computer interface. Nevertheless, few machine learning method can be used as the state-of-the-art method in the multi-class classification of limb movements. This work focuses on the multi-class classification of upper limb movements and proposes the multi-class filter bank task-related component analysis (mFBTRCA) method, which consists of three steps: spatial filtering, similarity measuring and filter bank selection. The spatial filter, namely the task-related component analysis, is first used to remove noise from EEG signals. The canonical correlation measures the similarity of the spatial-filtered signals and is used for feature extraction. The correlation features are extracted from multiple low-frequency filter banks. The minimum-redundancy maximum-relevance selects the essential features from all the correlation features, and finally, the support vector machine is used to classify the selected features. The proposed method compared against previously used models is evaluated using two datasets. mFBTRCA achieved a classification accuracy of 0.4193 ± 0.0780 (7 classes) and 0.4032 ± 0.0714 (5 classes), respectively, which improves on the best accuracies achieved using the compared methods (0.3590 ± 0.0645 and 0.3159 ± 0.0736, respectively). The proposed method is expected to provide more control commands in the applications of non-invasive brain-computer interfaces.

Classification of the Epileptic Seizure Onset Zone Based on Partial Annotation.

Epilepsy is a chronic disorder caused by excessive electrical discharges. Currently, clinical experts identify the seizure onset zone (SOZ) channel through visual judgment based on long-time intracranial electroencephalogram (iEEG), which is a very time-consuming, difficult and experience-based task. Therefore, there is a need for high-accuracy diagnostic aids to reduce the workload of clinical experts. In this article, we propose a method in which, the iEEG is split into the 20-s segment and for each patient, we ask clinical experts to label a part of the data, which is used to train a model and classify the remaining iEEG data. In recent years, machine learning methods have been successfully applied to solve some medical problems. Filtering, entropy and short-time Fourier transform (STFT) are used for extracting features. We compare them to wavelet transform (WT), empirical mode decomposition (EMD) and other traditional methods with the aim of obtaining the best possible discriminating features. Finally, we look for their medical interpretation, which is important for clinical experts. We achieve high-performance results for SOZ and non-SOZ data classification by using the labeled iEEG data and support vector machine (SVM), fully connected neural network (FCNN) and convolutional neural network (CNN) as classification models. In addition, we introduce the positive unlabeled (PU) learning to further reduce the workload of clinical experts. By using PU learning, we can learn a binary classifier with a small amount of labeled data and a large amount of unlabeled data. This can greatly reduce the amount and difficulty of annotation work by clinical experts. All together, we show that using 105 minutes of labeled data we achieve a classification result of 91.46% on average for multiple patients.

Fusion of EEG and Eye Blink Analysis for Detection of Driver Fatigue.

OBJECTIVE: The driver fatigue detection using multi-channel electroencephalography (EEG) has been extensively addressed in the literature. However, the employment of a single prefrontal EEG channel should be prioritized as it provides users with more comfort. Furthermore, eye blinks from such channel can be analyzed as the complementary information. Here, we present a new driver fatigue detection method based on simultaneous EEG and eye blinks analysis using an Fp1 EEG channel. METHODS: First, the moving standard deviation algorithm identifies eye blink intervals (EBIs) to extract blink-related features. Second, the discrete wavelet transform filters the EBIs from the EEG signal. Third, the filtered EEG signal is decomposed into sub-bands, and various linear and nonlinear features are extracted. Finally, the prominent features are selected by the neighbourhood components analysis and fed to a classifier to discriminate between fatigue and alert driving. In this paper, two different databases are investigated. The first one is used for parameters' tuning of proposed method for the eye blink detection and filtering, nonlinear EEG measures, and feature selection. The second one is solely used for testing the robustness of the tuned parameters. MAIN RESULTS: The comparison between the obtained results from both databases by the AdaBoost classifier in terms of sensitivity (90.2% vs. 87.4%), specificity (87.7% vs. 85.5%), and accuracy (88.4% vs. 86.8%) indicates the reliability of the proposed method for the driver fatigue detection. SIGNIFICANCE: Considering the existence of commercial single prefrontal channel EEG headbands, the proposed method can be used to detect the driver fatigue in real-world scenarios.

Knowledge and perception of primary care healthcare professionals on the use of artificial intelligence as a healthcare tool.

Objective: The rapid digitisation of healthcare data and the sheer volume being generated means that artificial intelligence (AI) is becoming a new reality in the practice of medicine. For this reason, describing the perception of primary care (PC) healthcare professionals on the use of AI as a healthcare tool and its impact in radiology is crucial to ensure its successful implementation. Methods: Observational cross-sectional study, using the validated Shinners Artificial Intelligence Perception survey, aimed at all PC medical and nursing professionals in the health region of Central Catalonia. Results: The survey was sent to 1068 health professionals, of whom 301 responded. And 85.7% indicated that they understood the concept of AI but there were discrepancies in the use of this tool; 65.8% indicated that they had not received any AI training and 91.4% that they would like to receive training. The mean score for the professional impact of AI was 3.62 points out of 5 (standard deviation (SD) = 0.72), with a higher score among practitioners who had some prior knowledge of and interest in AI. The mean score for preparedness for AI was 2.76 points out of 5 (SD = 0.70), with higher scores for nursing and those who use or do not know if they use AI. Conclusions: The results of this study show that the majority of professionals understood the concept of AI, perceived its impact positively, and felt prepared for its implementation. In addition, despite being limited to a diagnostic aid, the implementation of AI in radiology was a high priority for these professionals.

Seizure onset zone classification based on imbalanced iEEG with data augmentation.

Objective. Identifying the seizure onset zone (SOZ) in patients with focal epilepsy is the critical information required for surgery. However, collecting this information is challenging, time-consuming, and subjective. Some machine learning methods reduce the workload of clinical experts in intracranial electroencephalogram (iEEG) visual diagnosis but face significant challenges because interictal iEEG clinical data often suffer from a significant class imbalance. We aim to generate synthetic data for the minority class.Approach. To make the clinically imbalanced data suitable for machine learning, we introduce an EEG augmentation method (EEGAug). The EEGAug method randomly selects several samples from the minority class and transforms them into the frequency domain. Then, different frequency bands from different samples are used to compose new data. Finally, a synthetic sample is generated after converting the new data back to the time domain.Main results. The imbalanced clinical iEEG data can be balanced and applied to machine learning models using the method. A one-dimensional convolutional neural network model is used to classify the SOZ and non-SOZ data. We compare the EEGAug method with other data augmentation methods and another method of class-balanced focal loss function, which is also used for solving the data imbalance problem by adjusting the weights between the minority and majority classes. The results show that the EEGAug method performs best in most data.Significance. Data imbalance is a widespread clinical problem. The EEGAug method can flexibly generate synthetic data for the minority class, yielding synthetic and raw data with a high distribution similarity. By using the EEGAug method, clinical data can be used in machine learning models.