Non-linear Entropy Analysis in EEG to Predict Treatment Response to Repetitive Transcranial Magnetic Stimulation in Depression

Neurophysiological and neuroimaging markers of repetitive transcranial magnetic stimulation treatment response in major depressive disorder: A systematic review and meta-analysis of predictive modeling studies

Predicting repetitive transcranial magnetic stimulation (rTMS) treatment outcomes in major depressive disorder (MDD) could reduce the financial and psychological risks of treatment failure. We systematically reviewed and meta-analyzed studies that leveraged neurophysiological and neuroimaging markers to predict rTMS response in MDD. Five databases were searched from inception to May 25, 2023. The primary meta-analytic outcome was predictive accuracy pooled from classification models. Regression models were summarized qualitatively. A promising marker was identified if it showed a sensitivity and specificity of 80% or higher in at least two independent studies. Searching yielded 36 studies. Twenty-two classification modeling studies produced an estimated area under the summary receiver operating characteristic curve of 0.87 (95% CI = 0.83-0.92), with 86.8% sensitivity (95% CI = 80.6-91.2%) and 81.9% specificity (95% CI = 76.1-86.4%). Frontal theta cordance measured by electroencephalography is closest to proof of concept. Predicting rTMS response using neurophysiological and neuroimaging markers is promising for clinical decision-making. However, replications by different research groups are needed to establish rigorous markers.

GCTNet: a graph convolutional transformer network for major depressive disorder detection based on EEG signals

Objective.Identifying major depressive disorder (MDD) using objective physiological signals has become a pressing challenge.Approach.Hence, this paper proposes a graph convolutional transformer network (GCTNet) for accurate and reliable MDD detection using electroencephalogram (EEG) signals. The developed framework integrates a residual graph convolutional network block to capture spatial information and a Transformer block to extract global temporal dynamics. Additionally, we introduce the contrastive cross-entropy (CCE) loss that combines contrastive learning to enhance the stability and discriminability of the extracted features, thereby improving classification performance.Main results. The effectiveness of the GCTNet model and CCE loss was assessed using EEG data from 41 MDD patients and 44 normal controls, in addition to a publicly available dataset. Utilizing a subject-independent data partitioning method and 10-fold cross-validation, the proposed method demonstrated significant performance, achieving an average Area Under the Curve of 0.7693 and 0.9755 across both datasets, respectively. Comparative analyses demonstrated the superiority of the GCTNet framework with CCE loss over state-of-the-art algorithms in MDD detection tasks.Significance. The proposed method offers an objective and effective approach to MDD detection, providing valuable support for clinical-assisted diagnosis.

Graph-based Analysis to Predict Repetitive Transcranial Magnetic Stimulation Treatment Response in Patients With Major Depressive Disorder Using EEG Signals

Introduction

Repetitive transcranial magnetic stimulation (rTMS) is a non-pharmacological treatment for drug-resistant major depressive disorder (MDD) patients. Since the success rate of rTMS treatment is about 50%-55%, it is essential to predict the treatment outcome before starting based on electroencephalogram (EEG) signals, leading to identifying effective biomarkers and reducing the burden of health care centers.

Methods

To this end, pretreatment EEG data with 19 channels in the resting state from 34 drug-resistant MDD patients were recorded. Then, all patients received 20 sessions of rTMS treatment, and a reduction of at least 50% in the total beck depression inventory (BDI-II) score before and after the rTMS treatment was defined as a reference. In the current study, effective brain connectivity features were determined by the direct directed transfer function (dDTF) method from patients' pretreatment EEG data in all frequency bands separately. Then, the brain functional connectivity patterns were modeled as graphs by the dDTF method and examined with the local graph theory indices, including degree, out-degree, in-degree, strength, out-strength, in-strength, and betweenness centrality.

Results

The results indicated that the betweenness centrality index in the Fp2 node and the δ frequency band are the best biomarkers, with the highest area under the receiver operating characteristic curve value of 0.85 for predicting the rTMS treatment outcome in drug-resistant MDD patients.

Conclusion

The proposed method investigated the significant biomarkers that can be used to predict the rTMS treatment outcome in drug-resistant MDD patients and help clinical decisions.

Prediction of response to repetitive transcranial magnetic stimulation for major depressive disorder using hybrid Convolutional recurrent neural networks and raw Electroencephalogram Signal

Major Depressive Disorder (MDD) is a high prevalence disease that needs an effective and timely treatment to prevent its progress and additional costs. Repetitive Transcranial Magnetic Stimulation (rTMS) is an effective treatment option for MDD patients which uses strong magnetic pulses to stimulate specific regions of the brain. However, some patients do not respond to this treatment which causes the waste of multiple weeks as treatment time and clinical resources. Therefore developing an effective way for the prediction of response to the rTMS treatment of depression is necessary. In this work, we proposed a hybrid model created by pre-trained Convolutional Neural Networks (CNN) models and Bidirectional Long Short-Term Memory (BLSTM) cells to predict response to rTMS treatment from raw EEG signal. Three pre-trained CNN models named VGG16, InceptionResNetV2, and EffecientNetB0 were utilized as Transfer Learning (TL) models to construct hybrid TL-BLSTM models. Then an ensemble of these models was created using weighted majority voting which the weights were optimized by Differential Evolution (DE) optimization algorithm. Evaluation of these models shows the superior performance of the ensemble model by the accuracy of 98.51%, sensitivity of 98.64%, specificity of 98.36%, F1-score of 98.6%, and AUC of 98.5%. Therefore, the ensemble of the proposed hybrid convolutional recurrent networks can efficiently predict the treatment outcome of rTMS using raw EEG data.

A convolutional recurrent neural network with attention for response prediction to repetitive transcranial magnetic stimulation in major depressive disorder

Prediction of response to Repetitive Transcranial Magnetic Stimulation (rTMS) can build a very effective treatment platform that helps Major Depressive Disorder (MDD) patients to receive timely treatment. We proposed a deep learning model powered up by state-of-the-art methods to classify responders (R) and non-responders (NR) to rTMS treatment. Pre-treatment Electro-Encephalogram (EEG) signal of public TDBRAIN dataset and 46 proprietary MDD subjects were utilized to create time-frequency representations using Continuous Wavelet Transform (CWT) to be fed into the two powerful pre-trained Convolutional Neural Networks (CNN) named VGG16 and EfficientNetB0. Equipping these Transfer Learning (TL) models with Bidirectional Long Short-Term Memory (BLSTM) and attention mechanism for the extraction of most discriminative spatiotemporal features from input images, can lead to superior performance in the prediction of rTMS treatment outcome. Five brain regions named Frontal, Central, Parietal, Temporal, and occipital were assessed and the highest evaluated performance in 46 proprietary MDD subjects was acquired for the Frontal region using the TL-LSTM-Attention model based on EfficientNetB0 with accuracy, sensitivity, specificity, and Area Under the Curve (AUC) of 97.1%, 97.3%, 97.0%, and 0.96 respectively. Additionally, to test the generalizability of the proposed models, these TL-BLSTM-Attention models were evaluated on a public dataset called TDBRAIN and the highest accuracy of 82.3%, the sensitivity of 80.2%, the specificity of 81.9% and the AUC of 0.83 were obtained. Therefore, advanced deep learning methods using a time-frequency representation of EEG signals from the frontal brain region and the convolutional recurrent neural networks equipped with the attention mechanism can construct an accurate platform for the prediction of response to the rTMS treatment.

An effective brain connectivity technique to predict repetitive transcranial magnetic stimulation outcome for major depressive disorder patients using EEG signals

One of the most effective treatments for drug-resistant Major depressive disorder (MDD) patients is repetitive transcranial magnetic stimulation (rTMS). To improve treatment efficacy and reduce health care costs, it is necessary to predict the treatment response. In this study, we intend to predict the rTMS treatment response in MDD patients from electroencephalogram (EEG) signals before starting the treatment using machine learning approaches. Effective brain connectivity of 19-channel EEG data of MDD patients was calculated by the direct directed transfer function (dDTF) method. Then, using three feature selection methods, the best features were selected and patients were classified as responders or non-responders to rTMS treatment by using the support vector machine (SVM). Results on the 34 MDD patients indicated that the Fp2 region in the delta and theta frequency bands has a significant difference between the two groups and can be used as a significant brain biomarker to assess the rTMS treatment response. Also, the highest accuracy (89.6%) using the SVM classifier for the best features of the dDTF method based on the area under the receiver operating characteristic curve (AUC-ROC) criteria was obtained by combining the delta and theta frequency bands. Consequently, the proposed method can accurately detect the rTMS treatment response in MDD patients before starting treatment on the EEG signal to avoid financial and time costs to patients and medical centers.

Nonlinear Analyses Distinguish Load Carriage Dynamics in Walking and Standing: A Systematic Review

Load carriage experiments are typically performed from a linear perspective that assumes that movement variability is equivalent to error or noise in the neuromuscular system. A complimentary, nonlinear perspective that treats variability as the object of study has generated important results in movement science outside load carriage settings. To date, no systematic review has yet been conducted to understand how load carriage dynamics change from a nonlinear perspective. The goal of this systematic review is to fill that need. Relevant literature was extracted and reviewed for general trends involving nonlinear perspectives on load carriage. Nonlinear analyses that were used in the reviewed studies included sample, multiscale, and approximate entropy; the Lyapunov exponent; fractal analysis; and relative phase. In general, nonlinear tools successfully distinguish between unloaded and loaded conditions in standing and walking, although not in a consistent manner. The Lyapunov exponent and entropy were the most used nonlinear methods. Two noteworthy findings are that entropy in quiet standing studies tends to decrease, whereas the Lyapunov exponent in walking studies tends to increase, both due to added load. Thus, nonlinear analyses reveal altered load carriage dynamics, demonstrating promise in applying a nonlinear perspective to load carriage while also underscoring the need for more research.

Entropy and the Emotional Brain: Overview of a Research Field

During the last years, there has been a notable increase in the number of studies focused on the assessment of brain dynamics for the recognition of emotional states by means of nonlinear methodologies. More precisely, different entropy metrics have been applied for the analysis of electroencephalographic recordings for the detection of emotions. In this sense, regularity-based entropy metrics, symbolic predictability-based entropy indices, and different multiscale and multilag variants of the aforementioned methods have been successfully tested in a series of studies for emotion recognition from the EEG recording. This chapter aims to unify all those contributions to this scientific area, summarizing the main discoverings recently achieved in this research field.

An Optimal Channel Selection for EEG-Based Depression Detection via Kernel-Target Alignment

Depression is a mental disorder with emotional and cognitive dysfunction. The main clinical characteristic of depression is significant and persistent low mood. As reported, depression is a leading cause of disability worldwide. Moreover, the rate of recognition and treatment for depression is low. Therefore, the detection and treatment of depression are urgent. Multichannel electroencephalogram (EEG) signals, which reflect the working status of the human brain, can be used to develop an objective and promising tool for augmenting the clinical effects in the diagnosis and detection of depression. However, when a large number of EEG channels are acquired, the information redundancy and computational complexity of the EEG signals increase; thus, effective channel selection algorithms are required not only for machine learning feasibility, but also for practicality in clinical depression detection. Consequently, we propose an optimal channel selection method for EEG-based depression detection via kernel-target alignment (KTA) to effectively resolve the abovementioned issues. In this method, we consider a modified version KTA that can measure the similarity between the kernel matrix for channel selection and the target matrix as an objective function and optimize the objective function by a proposed optimal channel selection strategy. Experimental results on two EEG datasets show that channel selection can effectively increase the classification performance and that even if we rely only on a small subset of channels, the results are still acceptable. The selected channels are in line with the expected latent cortical activity patterns in depression detection. Moreover, the experimental results demonstrate that our method outperforms the state-of-the-art channel selection approaches.

Neural Correlates of Motor Recovery after Robot-Assisted Training in Chronic Stroke: A Multimodal Neuroimaging Study

Stroke is a leading cause of motor disability worldwide, and robot-assisted therapies have been increasingly applied to facilitate the recovery process. However, the underlying mechanism and induced neuroplasticity change remain partially understood, and few studies have investigated this from a multimodality neuroimaging perspective. The current study adopted BCI-guided robot hand therapy as the training intervention and combined multiple neuroimaging modalities to comprehensively understand the potential association between motor function alteration and various neural correlates. We adopted EEG-informed fMRI technique to understand the functional regions sensitive to training intervention. Additionally, correlation analysis among training effects, nonlinear property change quantified by fractal dimension (FD), and integrity of M1-M1 (M1: primary motor cortex) anatomical connection were performed. EEG-informed fMRI analysis indicated that for iM1 (iM1: ipsilesional M1) regressors, regions with significantly increased partial correlation were mainly located in contralesional parietal, prefrontal, and sensorimotor areas and regions with significantly decreased partial correlation were mainly observed in the ipsilesional supramarginal gyrus and superior temporal gyrus. Pearson's correlations revealed that the interhemispheric asymmetry change significantly correlated with the training effect as well as the integrity of M1-M1 anatomical connection. In summary, our study suggested that multiple functional brain regions not limited to motor areas were involved during the recovery process from multimodality perspective. The correlation analyses suggested the essential role of interhemispheric interaction in motor rehabilitation. Besides, the underlying structural substrate of the bilateral M1-M1 connection might relate to the interhemispheric change. This study might give some insights in understanding the neuroplasticity induced by the integrated BCI-guided robot hand training intervention and further facilitate the design of therapies for chronic stroke patients.

Personalization of Repetitive Transcranial Magnetic Stimulation for the Treatment of Major Depressive Disorder According to the Existing Psychiatric Comorbidity

Repetitive transcranial magnetic stimulation (rTMS) and intermittent theta-burst stimulation (iTBS) are evidenced-based treatments for patients with major depressive disorder (MDD) who fail to respond to standard first-line therapies. However, although various TMS protocols have been proven to be clinically effective, the response rate varies across clinical applications due to the heterogeneity of real-world psychiatric comorbidities, such as generalized anxiety disorder, posttraumatic stress disorder, panic disorder, or substance use disorder, which are often observed in patients with MDD. Therefore, individualized treatment approaches are important to increase treatment response by assigning a given patient to the most optimal TMS treatment protocol based on his or her individual profile. This literature review summarizes different rTMS or TBS protocols that have been applied in researches investigating MDD patients with certain psychiatric comorbidities and discusses biomarkers that may be used to predict rTMS treatment response. Furthermore, we highlight the need for the validation of neuroimaging and electrophysiological biomarkers associated with rTMS treatment responses. Finally, we discuss on which directions future efforts should focus for developing the personalization of the treatment of depression with rTMS or iTBS.

Analysis of functional MRI signal complexity based on permutation fuzzy entropy in bipolar disorder

Bipolar disorder is a manifestation of an emotional disease and is associated with emotional and cognitive dysfunction. The entropy-based method has been widely used to study the complexity of resting-state functional MRI (rs-fMRI) signals in mental diseases; however, alterations in the brain rs-fMRI signal complexities in bipolar disorder patients remain unclear, and previously used entropy methods are sensitive to noise. Here, we performed a work using permutation fuzzy entropy (PFEN), which has better performance than previously used methods, to analyze the brain complexity of bipolar disorder patients. Based on PFEN research, we obtained brain entropy maps of 49 bipolar disorder patients and 49 normal control, extracted the regions of interest to analyze the complexity of abnormal brain regions and further analyzed the correlation between the PFEN values of abnormal brain regions and the clinical measurement scores. Compared with the values in the normal control group, we found that significantly increased PFEN values mainly appeared in the middle temporal gyrus, angular gyrus, superior occipital gyrus and medial superior frontal gyrus, and the decreased PFEN values were found in the inferior temporal gyrus in bipolar disorder patients. In addition, the PFEN values of the angular gyrus was significantly negatively correlated with clinical scores. These findings improve our understanding of the pathophysiology of bipolar disorder patients.

The Changes of qEEG Approximate Entropy during Test of Variables of Attention as a Predictor of Major Depressive Disorder

Evaluating brain function through biosignals remains challenging. Quantitative electroencephalography (qEEG) outcomes have emerged as a potential intermediate biomarker for diagnostic clarification in psychological disorders. The Test of Variables of Attention (TOVA) was combined with qEEG to evaluate biomarkers such as absolute power, relative power, cordance, and approximate entropy from covariance matrix images to predict major depressive disorder (MDD). EEG data from 18 healthy control and 18 MDD patients were monitored during the resting state and TOVA. TOVA was found to provide aspects for the evaluation of MDD beyond resting electroencephalography. The results showed that the prefrontal qEEG theta cordance of the control and MDD groups were significantly different. For comparison, the changes in qEEG approximate entropy (ApEn) patterns observed during TOVA provided features to distinguish between participants with or without MDD. Moreover, ApEn scores during TOVA were a strong predictor of MDD, and the ApEn scores correlated with the Beck Depression Inventory (BDI) scores. Between-group differences in ApEn were more significant for the testing state than for the resting state. Our results provide further understanding for MDD treatment selection and response prediction during TOVA.

Entropy: A Promising EEG Biomarker Dichotomizing Subjects With Opioid Use Disorder and Healthy Controls

Electroencephalography (EEG) signals are known to be nonstationary and often multicomponential signals containing information about the condition of the brain. Since the EEG signal has complex, nonlinear, nonstationary, and highly random behaviour, numerous linear feature extraction methods related to the short-time windowing technique do not satisfy higher classification accuracy. Since biosignals are highly subjective, the symptoms may appear at random in the time scale and very small variations in EEG signals may depict a definite type of brain abnormality it is valuable and vital to extract and analyze the EEG signal parameters using computers. The challenge is to design and develop signal processing algorithms that extract this subtle information and use it for diagnosis, monitoring, and treatment of subjects suffering from psychiatric disorders. For this purpose, finite impulse response-based filtering process was employed rather than traditional time and frequency domain methods. Finite impulse response subbands were analyzed further to obtain feature vectors of different entropy markers and these features were fed into a classifier namely multilayer perceptron. The performances of the classifiers were finally compared considering overall classification accuracies, area under receiver operating characteristic curve scores. Our results underline the potential benefit of the introduced methodology is promising and is to be treated as a clinical interface in dichotomizing substance use disorders subjects and for other medical data analysis studies. The results also indicate that entropy estimators can distinguish normal and opioid use disorder subjects. EEG data and theta frequency band have distinctive capability for almost all types of entropies while nonextensive Tsallis entropy outperforms compared with other types of entropies.

Transient Dose-dependent Effects of Ketamine on Neural Oscillatory Activity in Wistar-Kyoto Rats

Ketamine is a promising therapeutic for treatment-resistant depression (TRD) but is associated with an array of short-term psychomimetic side-effects. These disparate drug effects may be elicited through the modulation of neural circuit activity. The purpose of this study was to therefore delineate dose- and time-dependent changes in ketamine-induced neural oscillatory patterns in regions of the brain implicated in depression. Wistar-Kyoto rats were used as a model system to study these aspects of TRD neuropathology whereas Wistar rats were used as a control strain. Animals received a low (10 mg/kg) or high (30 mg/kg) dose of ketamine and temporal changes in neural oscillatory activity recorded from the prefrontal cortex (PFC), cingulate cortex (Cg), and nucleus accumbens (NAc) for ninety minutes. Effects of each dose of ketamine on immobility in the forced swim test were also evaluated. High dose ketamine induced a transient increase in theta power in the PFC and Cg, as well as a dose-dependent increase in gamma power in these regions 10-min, but not 90-min, post-administration. In contrast, only low dose ketamine normalized innate deficits in fast gamma coherence between the NAc-Cg and PFC-Cg, an effect that persisted at 90-min post-injection. These low dose ketamine-induced oscillatory alterations were accompanied by a reduction in immobility time in the forced swim test. These results show that ketamine induces time-dependent effects on neural oscillations at specific frequencies. These drug-induced changes may differentially contribute to the psychomimetic and therapeutic effects of the drug.

Precision non-implantable neuromodulation therapies: a perspective for the depressed brain

Current first-line treatments for major depressive disorder (MDD) include pharmacotherapy and cognitive-behavioral therapy. However, one-third of depressed patients do not achieve remission after multiple medication trials, and psychotherapy can be costly and time-consuming. Although non-implantable neuromodulation (NIN) techniques such as transcranial magnetic stimulation, transcranial direct current stimulation, electroconvulsive therapy, and magnetic seizure therapy are gaining momentum for treating MDD, the efficacy of non-convulsive techniques is still modest, whereas use of convulsive modalities is limited by their cognitive side effects. In this context, we propose that NIN techniques could benefit from a precision-oriented approach. In this review, we discuss the challenges and opportunities in implementing such a framework, focusing on enhancing NIN effects via a combination of individualized cognitive interventions, using closed-loop approaches, identifying multimodal biomarkers, using computer electric field modeling to guide targeting and quantify dosage, and using machine learning algorithms to integrate data collected at multiple biological levels and identify clinical responders. Though promising, this framework is currently limited, as previous studies have employed small samples and did not sufficiently explore pathophysiological mechanisms associated with NIN response and side effects. Moreover, cost-effectiveness analyses have not been performed. Nevertheless, further advancements in clinical trials of NIN could shift the field toward a more “precision-oriented” practice.

Identification of Clinical Features and Biomarkers that may inform a Personalized Approach to rTMS for Depression

Repetitive transcranial magnetic stimulation (rTMS), an established treatment for treatment-resistant depression, may hold promise as a personalized medicine approach for the treatment of major depressive disorder (MDD). Clinical research has begun to identify patient-specific factors that could be used to guide rTMS treatment decisions or individualized treatment approaches. This literature review describes a range of patient factors which have been evaluated as potential biomarkers of rTMS treatment response, including patient- and illness-related characteristics, genetic factors, and biomarkers derived from neuroimaging and EEG. We highlight the need for validation data for imaging and electrophysiological biomarkers associated with rTMS as well as prospective evaluation of clinical predictors. Finally, we consider implications for future efforts to move toward a personalized medicine approach in the treatment of depression with rTMS.

Predictive Value of Slow and Fast EEG Oscillations for Methylphenidate Response in ADHD

Attention-deficit/hyperactivity disorder (ADHD) is the most common neurodevelopmental disorder and is characterized by symptoms of inattention and/or hyperactivity and impulsivity. In the current study, we obtained quantitative EEG (QEEG) recordings of 51 children aged between 6 and 12 years before the initiation of methylphenidate treatment. The relationship between changes in the scores of ADHD symptoms and initial QEEG features (power/power ratios values) were assessed. In addition, the children were classified as responder and nonresponder according to the ratio of their response to the medication (>25% improvement after medication). Logistic regression analyses were performed to analyze the accuracy of QEEG features for predicting responders. The findings indicate that patients with increased delta power at F8, theta power at Fz, F4, C3, Cz, T5, and gamma power at T6 and decreased beta powers at F8 and P3 showed more improvement in ADHD hyperactivity symptoms. In addition, increased delta/beta power ratio at F8 and theta/beta power ratio at F8, F3, Fz, F4, C3, Cz, P3, and T5 showed negative correlations with Conners' score difference of hyperactivity as well. This means, those with greater theta/beta and delta/beta powers showed more improvement in hyperactivity following medication. Theta power at Cz and T5 and theta/beta power ratios at C3, Cz, and T5 have significantly classified responders and nonresponders according to the logistic binary regression analysis. The results show that slow and fast oscillations may have predictive value for treatment response in ADHD. Future studies should seek for more sensitive biomarkers.

Intrinsic Brain Network Biomarkers of Antidepressant Response: a Review

PURPOSE OF REVIEW

Poor treatment response is a hallmark of major depressive disorder. To tackle this problem, recent neuroimaging studies have sought to characterize antidepressant response in terms of pretreatment differences in intrinsic functional brain networks. Our aim is to review recent studies that predict antidepressant response using intrinsic network connectivity. We discuss current methodological limitations and directions for future antidepressant biomarker studies.

RECENT FINDINGS

Functional connectivity stemming from the subgenual and rostral anterior cingulate has shown particular consistency in predicting antidepressant response. Differences in this connectivity may prove fruitful in differentiating treatment responders to many antidepressant interventions. Future biomarker studies should integrate biological MDD subtypes to address the disorder's inherent clinical heterogeneity. These clinical and scientific advancements have the potential to address this population marked by limited treatment response. Methodological considerations, including patient selection, response criteria, and model overfitting, will require future investigation to ensure that biomarkers generalize for prospective prediction of treatment response.