Unraveling nonlinear electrophysiologic processes in the human visual system with full dimension spectral analysis

Analysis of Amplitude Modulation of EEG Based on Holo-Hilbert Spectrum Analysis During General Anesthesia

OBJECTIVE

The study aims to investigate the relationship between amplitude modulation (AM) of EEG and anesthesia depth during general anesthesia.

METHODS

In this study, Holo-Hilbert spectrum analysis (HHSA) was used to decompose the multichannel EEG signals of 15 patients to obtain the spatial distribution of AM in the brain. Subsequently, HHSA was applied to the prefrontal EEG (Fp1) obtained during general anesthesia surgery in 15 and 34 patients, and the α-θ and α-δ regions of feature (ROFs) were defined in Holo-Hilbert spectrum (HHS) and three features were derived to quantify AM in ROFs.

RESULTS

During anesthetized phase, an anteriorization of the spatial distribution of AMs of α-carrier in brain was observed, as well as AMs of α-θ and α-δ in the EEG of Fp1. The total power ([Formula: see text]), mean carrier frequency ([Formula: see text]) and mean amplitude frequency ([Formula: see text]) of AMs changed during different anesthesia states.

CONCLUSION

HHSA can effectively analyze the cross-frequency coupling of EEG during anesthesia and the AM features may be applied to anesthesia monitoring.

SIGNIFICANCE

The study provides a new perspective for the characterization of brain states during general anesthesia, which is of great significance for exploring new features of anesthesia monitoring.

Order-frequency Holo-Hilbert spectral analysis for machinery fault diagnosis under time-varying operating conditions

Holo-Hilbert spectral analysis (HHSA) has been demonstrated to be an effective instantaneous feature demodulation tool for revealing the coupling relationship between the frequency-modulated (FM) carriers and amplitude-modulated (AM) characteristics within nonlinear and non-stationary mechanical vibration signals. However, it is unable to acquire the time varying AM characteristics from the vibration signals of the equipment operates under variable speed conditions. To decode such signals, inspired by HHSA, a novel angle-time double-layer decomposition structure termed order-frequency HHSA (OFHHSA) is established to demodulate the fault information from the time varying vibration signals in this paper. The corresponding spectrogram, namely, order-frequency Holo-Hilbert spectrum (OFHHS) is acquired for describing the interaction relationship between time and angle domains. Besides, the order AM-marginal spectrum is derived from the OFHHS via integrating the carrier variable to exhibit the fault characteristic-related orders. Moreover, the differences between OFHHSA and angle-time cyclo-stationary framework-based order-frequency spectral correlation (OFSC) are analyzed for time varying machinery fault diagnosis. Finally, from the analyses of simulated and tested data of mechanical equipment, the OFHHSA method has avoided the limitations of the two OFSC estimators on periodic assumption and the maximum cut-off order, and the proposed method obtained a more accurate rate in fault identification and more robust ability of anti-noise.

Burst c-VEP Based BCI: Optimizing stimulus design for enhanced classification with minimal calibration data and improved user experience

The utilization of aperiodic flickering visual stimuli under the form of code-modulated Visual Evoked Potentials (c-VEP) represents a pivotal advancement in the field of reactive Brain-Computer Interface (rBCI). A major advantage of the c-VEP approach is that the training of the model is independent of the number and complexity of targets, which helps reduce calibration time. Nevertheless, the existing designs of c-VEP stimuli can be further improved in terms of visual user experience but also to achieve a higher signal-to-noise ratio, while shortening the selection time and calibration process. In this study, we introduce an innovative variant of code-VEP, referred to as "Burst c-VEP". This original approach involves the presentation of short bursts of aperiodic visual flashes at a deliberately slow rate, typically ranging from two to four flashes per second. The rationale behind this design is to leverage the sensitivity of the primary visual cortex to transient changes in low-level stimuli features to reliably elicit distinctive series of visual evoked potentials. In comparison to other types of faster-paced code sequences, burst c-VEP exhibit favorable properties to achieve high bitwise decoding performance using convolutional neural networks (CNN), which yields potential to attain faster selection time with the need for less calibration data. Furthermore, our investigation focuses on reducing the perceptual saliency of c-VEP through the attenuation of visual stimuli contrast and intensity to significantly improve users' visual comfort. The proposed solutions were tested through an offline 4-classes c-VEP protocol involving 12 participants. Following a factorial design, participants were instructed to focus on c-VEP targets whose pattern (burst and maximum-length sequences) and amplitude (100% or 40% amplitude depth modulations) were manipulated across experimental conditions. Firstly, the full amplitude burst c-VEP sequences exhibited higher accuracy, ranging from 90.5% (with 17.6s of calibration data) to 95.6% (with 52.8s of calibration data), compared to its m-sequence counterpart (71.4% to 85.0%). The mean selection time for both types of codes (1.5 s) compared favorably to reports from previous studies. Secondly, our findings revealed that lowering the intensity of the stimuli only slightly decreased the accuracy of the burst code sequences to 94.2% while leading to substantial improvements in terms of user experience. Taken together, these results demonstrate the high potential of the proposed burst codes to advance reactive BCI both in terms of performance and usability. The collected dataset, along with the proposed CNN architecture implementation, are shared through open-access repositories.

Single patient learning for adaptive radiotherapy dose prediction

BACKGROUND

Throughout a patient's course of radiation therapy, maintaining accuracy of their initial treatment plan over time is challenging due to anatomical changes-for example, stemming from patient weight loss or tumor shrinkage. Online adaptation of their RT plan to these changes is crucial, but hindered by manual and time-consuming processes. While deep learning (DL) based solutions have shown promise in streamlining adaptive radiation therapy (ART) workflows, they often require large and extensive datasets to train population-based models.

PURPOSE

This study extends our prior research by introducing a minimalist approach to patient-specific adaptive dose prediction. In contrast to our prior method, which involved fine-tuning a pre-trained population model, this new method trains a model from scratch using only a patient's initial treatment data. This patient-specific dose predictor aims to enhance clinical accessibility, thereby empowering physicians and treatment planners to make more informed, quantitative decisions in ART. We hypothesize that patient-specific DL models will provide more accurate adaptive dose predictions for their respective patients compared to a population-based DL model.

METHODS

We selected 33 patients to train an adaptive population-based (AP) model. Ten additional patients were selected, and their respective initial RT data served as single samples for training patient-specific (PS) models. These 10 patients contained an additional 26 ART plans that were withheld as the test dataset to evaluate AP versus PS model dose prediction performance. We assessed model performance using Mean Absolute Percent Error (MAPE) by comparing predicted doses to the originally delivered ground truth doses. We used the Wilcoxon signed-rank test to determine statistically significant differences in terms of MAPE between the AP and PS model results across the test dataset. Furthermore, we calculated differences between predicted and ground truth mean doses for segmented structures and determined statistical significance in the differences for each of them.

RESULTS

The average MAPE across AP and PS model dose predictions was 5.759% and 4.069%, respectively. The Wilcoxon signed-rank test yielded two-tailed p-value = 2.9802 × 10 - 8 $2.9802\ \times \ {10}^{ - 8}$ , indicating that the MAPE differences between the AP and PS model dose predictions are statistically significant, and 95% confidence interval = [-2.1610, -1.0130], indicating 95% confidence that the MAPE difference between the AP and PS models for a population lies in this range. Out of 24 total segmented structures, the comparison of mean dose differences for 12 structures indicated statistical significance with two-tailed p-values < 0.05.

CONCLUSION

Our study demonstrates the potential of patient-specific deep learning models in application to ART. Notably, our method streamlines the training process by minimizing the size of the required training dataset, as only a single patient's initial treatment data is required. External institutions considering the implementation of such a technology could package such a model so that it only requires the upload of a reference treatment plan for model training and deployment. Our single patient learning strategy demonstrates promise in ART due to its minimal dataset requirement and its utility in personalization of cancer treatment.

GPU Implementation of the Improved CEEMDAN Algorithm for Fast and Efficient EEG Time-Frequency Analysis

Time-frequency analysis of EEG data is a key step in exploring the internal activities of the human brain. Studying oscillations is an important part of the analysis, as they are thought to provide the underlying mechanism for communication between neural assemblies. Traditional methods of analysis, such as Short-Time FFT and Wavelet Transforms, are not ideal for this task due to the time-frequency uncertainty principle and their reliance on predefined basis functions. Empirical Mode Decomposition and its variants are more suited to this task as they are able to extract the instantaneous frequency and phase information but are too time consuming for practical use. Our aim was to design and develop a massively parallel and performance-optimized GPU implementation of the Improved Complete Ensemble EMD with the Adaptive Noise (CEEMDAN) algorithm that significantly reduces the computational time (from hours to seconds) of such analysis. The resulting GPU program, which is publicly available, was validated against a MATLAB reference implementation and reached over a 260× speedup for actual EEG measurement data, and provided predicted speedups in the range of 3000-8300× for longer measurements when sufficient memory was available. The significance of our research is that this implementation can enable researchers to perform EMD-based EEG analysis routinely, even for high-density EEG measurements. The program is suitable for execution on desktop, cloud, and supercomputer systems and can be the starting point for future large-scale multi-GPU implementations.

Automated treatment planning for proton pencil beam scanning using deep learning dose prediction and dose-mimicking optimization

PURPOSE

The purpose of this study is to investigate the use of a deep learning architecture for automated treatment planning for proton pencil beam scanning (PBS).

METHODS

A 3-dimensional (3D) U-Net model has been implemented in a commercial treatment planning system (TPS) that uses contoured regions of interest (ROI) binary masks as model inputs with a predicted dose distribution as the model output. Predicted dose distributions were converted to deliverable PBS treatment plans using a voxel-wise robust dose mimicking optimization algorithm. This model was leveraged to generate machine learning (ML) optimized plans for patients receiving proton PBS irradiation of the chest wall. Model training was carried out on a retrospective set of 48 previously-treated chest wall patient treatment plans. Model evaluation was carried out by generating ML-optimized plans on a hold-out set of 12 contoured chest wall patient CT datasets from previously treated patients. Clinical goal criteria and gamma analysis were used to compare dose distributions of the ML-optimized plans against the clinically approved plans across the test patients.

RESULTS

Statistical analysis of mean clinical goal criteria indicates that compared to the clinical plans, the ML optimization workflow generated robust plans with similar dose to the heart, lungs, and esophagus while achieving superior dosimetric coverage to the PTV chest wall (clinical mean V95 = 97.6% vs. ML mean V95 = 99.1%, p < 0.001) across the 12 test patients.

CONCLUSIONS

ML-based automated treatment plan optimization using the 3D U-Net model can generate treatment plans of similar clinical quality compared to human-driven optimization.

The association between working memory precision and the nonlinear dynamics of frontal and parieto-occipital EEG activity

Electrophysiological working memory (WM) research shows brain areas communicate via macroscopic oscillations across frequency bands, generating nonlinear amplitude modulation (AM) in the signal. Traditionally, AM is expressed as the coupling strength between the signal and a prespecified modulator at a lower frequency. Therefore, the idea of AM and coupling cannot be studied separately. In this study, 33 participants completed a color recall task while their brain activity was recorded through EEG. The AM of the EEG data was extracted using the Holo-Hilbert spectral analysis (HHSA), an adaptive method based on the Hilbert-Huang transforms. The results showed that WM load modulated parieto-occipital alpha/beta power suppression. Furthermore, individuals with higher frontal theta power and lower parieto-occipital alpha/beta power exhibited superior WM precision. In addition, the AM of parieto-occipital alpha/beta power predicted WM precision after presenting a target-defining probe array. The phase-amplitude coupling (PAC) between the frontal theta phase and parieto-occipital alpha/beta AM increased with WM load while processing incoming stimuli, but the PAC itself did not predict the subsequent recall performance. These results suggest frontal and parieto-occipital regions communicate through theta-alpha/beta PAC. However, the overall recall precision depends on the alpha/beta AM following the onset of the retro cue.

A holo-spectral EEG analysis provides an early detection of cognitive decline and predicts the progression to Alzheimer's disease

Aims

Our aim was to differentiate patients with mild cognitive impairment (MCI) and Alzheimer's disease (AD) from cognitively normal (CN) individuals and predict the progression from MCI to AD within a 3-year longitudinal follow-up. A newly developed Holo-Hilbert Spectral Analysis (HHSA) was applied to resting state EEG (rsEEG), and features were extracted and subjected to machine learning algorithms.

Methods

A total of 205 participants were recruited from three hospitals, with CN (n = 51, MMSE > 26), MCI (n = 42, CDR = 0.5, MMSE ≥ 25), AD1 (n = 61, CDR = 1, MMSE < 25), AD2 (n = 35, CDR = 2, MMSE < 16), and AD3 (n = 16, CDR = 3, MMSE < 16). rsEEG was also acquired from all subjects. Seventy-two MCI patients (CDR = 0.5) were longitudinally followed up with two rsEEG recordings within 3 years and further subdivided into an MCI-stable group (MCI-S, n = 36) and an MCI-converted group (MCI-C, n = 36). The HHSA was then applied to the rsEEG data, and features were extracted and subjected to machine-learning algorithms.

Results

(a) At the group level analysis, the HHSA contrast of MCI and different stages of AD showed augmented amplitude modulation (AM) power of lower-frequency oscillations (LFO; delta and theta bands) with attenuated AM power of higher-frequency oscillations (HFO; beta and gamma bands) compared with cognitively normal elderly controls. The alpha frequency oscillation showed augmented AM power across MCI to AD1 with a reverse trend at AD2. (b) At the individual level of cross-sectional analysis, implementation of machine learning algorithms discriminated between groups with good sensitivity (Sen) and specificity (Spec) as follows: CN elderly vs. MCI: 0.82 (Sen)/0.80 (Spec), CN vs. AD1: 0.94 (Sen)/0.80 (Spec), CN vs. AD2: 0.93 (Sen)/0.90 (Spec), and CN vs. AD3: 0.75 (Sen)/1.00 (Spec). (c) In the longitudinal MCI follow-up, the initial contrasted HHSA between MCI-S and MCI-C groups showed significantly attenuated AM power of alpha and beta band oscillations. (d) At the individual level analysis of longitudinal MCI groups, deploying machine learning algorithms with the best seven features resulted in a sensitivity of 0.9 by the support vector machine (SVM) classifier, with a specificity of 0.8 yielded by the decision tree classifier.

Conclusion

Integrating HHSA into EEG signals and machine learning algorithms can differentiate between CN and MCI as well as also predict AD progression at the MCI stage.

A review of critical brain oscillations in depression and the efficacy of transcranial magnetic stimulation treatment

Repetitive transcranial magnetic stimulation (rTMS) and intermittent theta burst stimulation (iTBS) have been proven effective non-invasive treatments for patients with drug-resistant major depressive disorder (MDD). However, some depressed patients do not respond to these treatments. Therefore, the investigation of reliable and valid brain oscillations as potential indices for facilitating the precision of diagnosis and treatment protocols has become a critical issue. The current review focuses on brain oscillations that, mostly based on EEG power analysis and connectivity, distinguish between MDD and controls, responders and non-responders, and potential depression severity indices, prognostic indicators, and potential biomarkers for rTMS or iTBS treatment. The possible roles of each biomarker and the potential reasons for heterogeneous results are discussed, and the directions of future studies are proposed.

Resting state dynamics in people with varying degrees of anxiety and mindfulness: A nonlinear and nonstationary perspective

Anxiety and mindfulness are two inversely linked traits shown to be involved in various physiological domains. The current study used resting state electroencephalography (EEG) to explore differences between people with low mindfulness-high anxiety (LMHA) (n = 29) and high mindfulness-low anxiety (HMLA) (n = 27). The resting EEG was collected for a total of 6 min, with a randomized sequence of eyes closed and eyes opened conditions. Two advanced EEG analysis methods, Holo-Hilbert Spectral Analysis and Holo-Hilbert cross-frequency phase clustering (HHCFPC) were employed to estimate the power-based amplitude modulation of carrier frequencies, and cross-frequency coupling between low and high frequencies, respectively. The presence of higher oscillation power across the delta and theta frequencies in the LMHA group than the HMLA group might have been due to the similarity between the resting state and situations of uncertainty, which reportedly triggers motivational and emotional arousal. Although these two groups were formed based on their trait anxiety and trait mindfulness scores, it was anxiety that was found to be significant predictor of the EEG power, not mindfulness. It led us to conclude that it might be anxiety, not mindfulness, which might have contributed to higher electrophysiological arousal. Additionally, a higher δ-β and δ-γ CFC in LMHA suggested greater local-global neural integration, consequently a greater functional association between cortex and limbic system than in the HMLA group. The present cross-sectional study may guide future longitudinal studies on anxiety aiming with interventions such as mindfulness to characterize the individuals based on their resting state physiology.

Three-dimensional dose and LETD prediction in proton therapy using artificial neural networks

PURPOSE

Challenges in proton therapy include identifying patients most likely to benefit; ensuring consistent, high-quality plans as its adoption becomes more widespread; and recognizing biological uncertainties that may be related to increased relative biologic effectiveness driven by linear energy transfer (LET). Knowledge-based planning (KBP) is a domain that may help to address all three.

METHODS

Artificial neural networks were trained using 117 unique treatment plans and associated dose and dose-weighted LET (LET_D ) distributions. The data set was split into training (n = 82), validation (n = 17), and test (n = 18) sets. Model performance was evaluated on the test set using dose- and LET_D -volume metrics in the clinical target volume (CTV) and nearby organs at risk and Dice similarity coefficients (DSC) comparing predicted and planned isodose lines at 50%, 75%, and 95% of the prescription dose.

RESULTS

Dose-volume metrics significantly differed (α = 0.05) between predicted and planned dose distributions in only one dose-volume metric, D_2% to the CTV. The maximum observed root mean square (RMS) difference between corresponding metrics was 4.3 Gy_RBE (8% of prescription) for D_1cc to optic chiasm. DSC were 0.90, 0.93, and 0.88 for the 50%, 75%, and 95% isodose lines, respectively. LET_D -volume metrics significantly differed in all but one metric, L_0.1cc of the brainstem. The maximum observed difference in RMS differences for LET_D metrics was 1.0 keV/μm for L_0.1cc to brainstem.

CONCLUSIONS

We have devised the first three-dimensional dose and LET_D -prediction model for cranial proton radiation therapy has been developed. Dose accuracy compared favorably with that of previously published models in other treatment sites. The agreement in LET_D supports future investigations with biological doses in mind to enable the full potential of KBP in proton therapy.

Amplitude modulating frequency overrides carrier frequency in tACS-induced phosphene percept

The amplitude modulated (AM) neural oscillation is an essential feature of neural dynamics to coordinate distant brain areas. The AM transcranial alternating current stimulation (tACS) has recently been adopted to examine various cognitive functions, but its neural mechanism remains unclear. The current study utilized the phosphene phenomenon to investigate whether, in an AM-tACS, the AM frequency could modulate or even override the carrier frequency in phosphene percept. We measured the phosphene threshold and the perceived flash rate/pattern from 12 human subjects (four females, aged from 20-44 years old) under tACS that paired carrier waves (10, 14, 18, 22 Hz) with different envelope conditions (0, 2, 4 Hz) over the mid-occipital and left facial areas. We also examined the phosphene source by adopting a high-density stimulation montage. Our results revealed that (1) phosphene threshold was higher for AM-tACS than sinusoidal tACS and demonstrated different carrier frequency functions in two stimulation montages. (2) AM-tACS slowed down the phosphene flashing and abolished the relation between the carrier frequency and flash percept in sinusoidal tACS. This effect was independent of the intensity change of the stimulation. (3) Left facial stimulation elicited phosphene in the upper-left visual field, while occipital stimulation elicited equally distributed phosphene. (4) The near-eye electrodermal activity (EDA) measured under the threshold-level occipital tACS was greater than the lowest power sufficient to elicit retinal phosphene. Our results show that AM frequency may override the carrier frequency and determine the perceived flashing frequency of AM-tACS-induced phosphene.

FB-EEGNet: A fusion neural network across multi-stimulus for SSVEP target detection

BACKGROUND

Steady-state visual evoked potential (SSVEP) is a prevalent paradigm of brain-computer interface (BCI). Recently, deep neural networks (DNNs) have been employed for SSVEP target recognition. However, current DNN models can not fully extract information from SSVEP harmonic components, and ignore the influence of non-target stimuli.

NEW METHOD

To employ information of multiple sub-bands and non-target stimulus data, we propose a DNN model for SSVEP target detection, i.e., FB-EEGNet, which fuses features of multiple neural networks. Additionally, we design a multi-label for each sample and optimize the parameters of FB-EEGNet across multi-stimulus to incorporate the information from non-target stimuli.

RESULTS

Under the subject-specific condition, FB-EEGNet achieves the average classification accuracies (information transfer rate (ITR)) of 76.75 % (50.70 bits/min) and 89.14 % (70.45 bits/min) in a time widow of 0.7 s under the public 12-target dataset and our experimental 9-target dataset, respectively. Under the cross-subject condition, FB-EEGNet achieved mean accuracies (ITRs) of 81.72 % (67.99 bits/min) and 92.15 % (76.12 bits/min) on the public and experimental datasets in a time window of 1 s, respectively.

COMPARISON WITH EXISTING METHODS

FB-EEGNet shows superior performance than CCNN, EEGNet, CCA and FBCCA both for subject-dependent and subject-independent SSVEP target recognition.

CONCLUSION

FB-EEGNet can effectively extract information from multiple sub-bands and cross-stimulus targets, providing a promising way for extracting deep features in SSVEP using neural networks.

Evaluating the Different Stages of Parkinson's Disease Using Electroencephalography With Holo-Hilbert Spectral Analysis

Electroencephalography (EEG) can reveal the abnormalities of dopaminergic subcortico-cortical circuits in patients with Parkinson's disease (PD). However, conventional time-frequency analysis of EEG signals cannot fully reveal the non-linear processes of neural activities and interactions. A novel Holo-Hilbert Spectral Analysis (HHSA) was applied to reveal non-linear features of resting state EEG in 99 PD patients and 59 healthy controls (HCs). PD patients demonstrated a reduction of β bands in frontal and central regions, and reduction of γ bands in central, parietal, and temporal regions. Compared with early-stage PD patients, late-stage PD patients demonstrated reduction of β bands in the posterior central region, and increased θ and δ2 bands in the left parietal region. θ and β bands in all brain regions were positively correlated with Hamilton depression rating scale scores. Machine learning algorithms using three prioritized HHSA features demonstrated "Bag" with the best accuracy of 0.90, followed by "LogitBoost" with an accuracy of 0.89. Our findings strengthen the application of HHSA to reveal high-dimensional frequency features in EEG signals of PD patients. The EEG characteristics extracted by HHSA are important markers for the identification of depression severity and diagnosis of PD.

Investigating the effect of Mindfulness-Based Stress Reduction on stress level and brain activity of college students

Financial constraints usually hinder students, especially those in low-middle income countries (LMICs), from seeking mental health interventions. Hence, it is necessary to identify effective, affordable and sustainable counter-stress measures for college students in the LMICs context. This study examines the sustained effects of mindfulness practice on the psychological outcomes and brain activity of students, especially when they are exposed to stressful situations. Here, we combined psychological and electrophysiological methods (EEG) to investigate the sustained effects of an 8-week-long standardized Mindfulness-Based Stress Reduction (MBSR) intervention on the brain activity of college students. We found that the Test group showed a decrease in negative emotional states after the intervention, compared to the no statistically significant result of the Control group, as indicated by the Perceived Stress Scale (PSS) (33% reduction in the negative score) and Depression, Anxiety, Stress Scale (DASS-42) scores (nearly 40% reduction of three subscale scores). Spectral analysis of EEG data showed that this intervention is longitudinally associated with increased frontal and occipital lobe alpha band power. Additionally, the increase in alpha power is more prevalent when the Test group was being stress-induced by cognitive tasks, suggesting that practicing MBSR might enhance the practitioners’ tolerance of negative emotional states. In conclusion, MBSR intervention led to a sustained reduction of negative emotional states as measured by both psychological and electrophysiological metrics, which supports the adoption of MBSR as an effective and sustainable stress-countering approach for students in LMICs.

Site-agnostic 3D dose distribution prediction with deep learning neural networks

PURPOSE

Typically, the current dose prediction models are limited to small amounts of data and require retraining for a specific site, often leading to suboptimal performance. We propose a site-agnostic, three-dimensional dose distribution prediction model using deep learning that can leverage data from any treatment site, thus increasing the total data available to train the model. Applying our proposed model to a new target treatment site requires only a brief fine-tuning of the model to the new data and involves no modifications to the model input channels or its parameters. Thus, it can be efficiently adapted to a different treatment site, even with a small training dataset.

METHODS

This study uses two separate datasets/treatment sites: data from patients with prostate cancer treated with intensity-modulated radiation therapy (source data), and data from patients with head-and-neck cancer treated with volumetric-modulated arc therapy (target data). We first developed a source model with 3D UNet architecture, trained from random initial weights on the source data. We evaluated the performance of this model on the source data. We then studied the generalizability of the model to the new target dataset via transfer learning. To do this, we built three more models, all with the same 3D UNet architecture: target model, adapted model, and combined model. The source and target models were trained on the source and target data from random initial weights, respectively. The adapted model fine-tuned the source model to the target domain by using the target data. Finally, the combined model was trained from random initial weights on a combined data pool consisting of both target and source datasets. We tested all four models on the target dataset and evaluated quantitative dose-volume histogram metrics for the planning target volume (PTV) and organs at risk (OARs).

RESULTS

When tested on the source treatment site, the source model accurately predicted the dose distributions with average (mean, max) absolute dose errors of (0.32%±0.14, 2.37%±0.93) (PTV) relative to the prescription dose, and highest mean dose error of 1.68%±0.76, and highest max dose error of 5.47%± 3.31 for femoral head right. The error in PTV dose coverage prediction is 3.21%±1.51 for D₉₈ , 3.04%±1.69 for D₉₅ , and 1.83%±1.01 for D₀₂ . Averaging across all OARs, the source model predicted the OAR mean dose within 1.38% and the OAR max dose within 3.64%. For the target treatment site, the target model average (mean, max) absolute dose errors relative to the prescription dose for the PTV were (1.08%±0.95, 2.90%±1.35). Left cochlea had the highest mean and max dose errors of 5.37%±5.82 and 8.33%±8.88, respectively. The errors in PTV dose coverage prediction for D₉₈ and D₉₅ were 2.88%±1.59 and 2.55%±1.28, respectively. The target model can predict the OAR mean dose within 2.43% and the OAR max dose within 4.33% on average across all OARs.

CONCLUSION

We developed a site-agnostic model for three-dimensional dose prediction and tested its adaptability to a new target treatment site via transfer learning. Our proposed model can make accurate predictions with limited training data.

Volumetric dose extension for isodose tuning

PURPOSE

To develop a method that can extend dose from two isodose surfaces (isosurfaces) to the entire patient volume, and to demonstrate its application in radiotherapy plan isodose tuning.

METHODS

We hypothesized that volumetric dose distribution can be extended from two isosurfaces-the 100% isosurface and a reference isosurface--with the distances to these two surfaces (L_ref ) as extension variables. The extension function is modeled by a three-dimensional lookup table (LUT), where voxel dose values from clinical plans are binned by three indexes: L₁₀₀ , L_ref and D_ref (reference dose level). The mean and standard deviation of voxel doses in each bin are calculated and stored in LUT. Volumetric dose extension is performed voxelwisely by indexing the LUT with the L₁₀₀ , L_ref and D_ref of each query voxel. The mean dose stored in the corresponding bin is filled into the query voxel as extended dose, and the standard deviation be filled voxelwisely as the uncertainty of extension result. We applied dose extension in isodose tuning, which aims to tune volumetric dose distribution by isosurface dragging. We adopted extended dose as an approximate dose estimation, and combined it with dose correction strategy to achieve accurate dose tuning.

RESULTS

We collected 32 post-operative prostate volumetric modulated arc therapy (VMAT) cases and built the LUT and its associated uncertainties from the doses of 27 cases. The dose extension method was tested on five cases, whose dose distributions were defined as ground truth (GT). We extended the doses from 100% and 50% GT isosurfaces to the entire volume, and evaluated the accuracy of extended doses. The 5 mm/5% gamma passing rate (GPR) of extended doses are 92.0%. The mean error is 4.5%, which is consistent to the uncertainty estimated by LUT. The dose difference in 90.5% of voxels is within two sigma and 97.5% in three sigma. The calculation time is less than two seconds. To simulate plan isodose tuning, we optimized a dose with less sparing on rectum (than GT dose) and defined it as a "base dose"-the dose awaiting isosurface dragging. In front-end, the simulated isodose tuning is conducted as such that the base dose was given to plan tuner, and its 50% isosurface would be dragged to the desired position (position of 50% isosurface in GT dose). In back-end, the output of isodose tuning is obtained by 1) extending dose from the desired isosurfaces and viewed the extended dose as an approximate dose, 2) obtaining a correction map from the base dose, and 3) applying the correction map to the extended dose. The accuracy of output-extended dose with correction-was 97.2% in GPR (3 mm/3%) and less than 1% in mean dose difference. The total calculation time is less than two seconds, which allows for interactive isodose tuning.

CONCLUSIONS

We developed a dose extension method that generates volumetric dose distribution from two surfaces. The application of dose extension is in interactive isodose tuning. The distance-based LUT fashion and correction strategy guarantee the computation efficiency and accuracy in isodose tuning. This article is protected by copyright. All rights reserved.

Multiscale Temporal Variations of Atmospheric Mercury Distinguished by the Hilbert-Huang Transform Analysis Reveals Multiple El Niño-Southern Oscillation Links

Atmospheric mercury (Hg) cycling is sensitive to climate-driven changes, but links with various teleconnections remain unestablished. Here, we revealed the El Niño-Southern Oscillation (ENSO) influence on gaseous elemental mercury (GEM) concentrations recorded at a background station in East Asia using the Hilbert-Huang transform (HHT). The timing and magnitude of GEM intrinsic variations were clearly distinguished by ensemble empirical mode decomposition (EEMD), revealing the amplitude of the GEM concentration interannual variability (IAV) is greater than that for diurnal and seasonal variability. We show that changes in the annual cycle of GEM were modulated by significant IAVs at time scales of 2-7 years, highlighted by a robust GEM IAV-ENSO relationship of the associated intrinsic mode functions. With confirmation that ENSO modulates the GEM annual cycle, we then found that weaker GEM annual cycles may have resulted from El Niño-accelerated Hg evasion from the ocean. Furthermore, the relationship between ENSO and GEM is sensitive to extreme events (i.e., 2015-2016 El Niño), resulting in perturbation of the long-term trend and atmospheric Hg cycling. Future climate change will likely increase the number of extreme El Niño events and, hence, could alter atmospheric Hg cycling and influence the effectiveness evaluation of the Minamata Convention on Mercury.

Holo-Hilbert spectral-based noise removal method for EEG high-frequency bands

Simultaneous EEG-fMRI is a growing and promising field, as it has great potential to further our understanding of the spatiotemporal dynamics of brain function in health and disease. In particular, there is much interest in understanding the fMRI correlates of brain activity in the gamma band (> 30 Hz), as these frequencies are thought to be associated with cognitive processes involving perception, attention, and memory, as well as with disorders such as schizophrenia and autism. However, progress in this area has been limited due to issues such as MR-induced artifacts in EEG recordings, which seem to be more problematic for gamma frequencies. This paper presents a noise removal method for the gamma band of EEG that is based on the Holo-Hilbert spectral analysis (HHSA), but with a new implementation strategy. HHSA uses a nested empirical mode decomposition (EMD) to identify amplitude and frequency modulations (AM and FM, respectively) by averaging over frequencies with high and significant powers. Our method examines gamma band by applying two layers of EMD to the FM and AM components, removing components with very low power based on the power-instantaneous frequency spectrum, and subsequently reconstructs the denoised gamma-band signal from the remaining components. Simulations demonstrate that our proposed method efficiently reduces artifacts while preserving the original gamma signal which is especially critical for simultaneous EEG/fMRI studies.

Critical role of rhythms in prefrontal transcranial magnetic stimulation for depression: A randomized sham-controlled study

Repetitive transcranial magnetic stimulation (rTMS) is an alternative treatment for depression, but the neural correlates of the treatment are currently inconclusive, which might be a limit of conventional analytical methods. The present study aimed to investigate the neurophysiological evidence and potential biomarkers for rTMS and intermittent theta burst stimulation (iTBS) treatment. A total of 61 treatment‐resistant depression patients were randomly assigned to receive prolonged iTBS (piTBS; N = 19), 10 Hz rTMS (N = 20), or sham stimulation (N = 22). Each participant went through a treatment phase with resting state electroencephalography (EEG) recordings before and after the treatment phase. The aftereffects of stimulation showed that theta‐alpha amplitude modulation frequency (f_am) was associated with piTBS_Responder, which involves repetitive bursts delivered in the theta frequency range, whereas alpha carrier frequency (f_c) was related to 10 Hz rTMS, which uses alpha rhythmic stimulation. In addition, theta‐alpha amplitude modulation frequency was positively correlated with piTBS antidepressant efficacy, whereas the alpha frequency was not associated with the 10 Hz rTMS clinical outcome. The present study showed that TMS stimulation effects might be lasting, with changes of brain oscillations associated with the delivered frequency. Additionally, theta‐alpha amplitude modulation frequency may be as a function of the degree of recovery in TRD with piTBS treatment and also a potential EEG‐based predictor of antidepressant efficacy of piTBS in the early treatment stage, that is, first 2 weeks.

Metformin improves epigenetic modification involved in oocyte growth and embryo development in polycystic ovary syndrome mice model

The possible relationship between dehydroepiandrosterone (DHEA)-induced polycystic ovary syndrome (PCOS) and epigenetic changes (ECs) leading to the impaired oocyte quality, has not been investigated yet. So, this study aimed to provide an insight into the relationship of the impaired oocyte quality with ECs in a mice DHEA-induced PCOS model and to further reveal the effect of metformin treatment. For this purpose, 80 female BALB/C mice were randomly divided into four equal groups, named as the control, sham, (DHEA) and DHEA + Metformin groups. The alterations in acetylation of H4K5 and H4K16, and in methylation of DNA (5MeC) and H3K9 were evaluated using immunocytochemical. Moreover, the expression of Hdac1, Hdac2, Dnmt1, and Dnmt3a genes involved in ECs were analyzed using reverse-transcription polymerase chain reaction. As well, the levels of mitochondrial membrane potential (MMP), oxidative stress (OS), embryo development, ovarian morphology, sexual hormone, ovulatory function, and AMPKα phosphorylation activity were compared in all the studied groups. Metformin attenuated the damages induced by DHEA as indicated by the normalized the estrous cycle, the improved ovarian morphology, the decreased sexual hormone and OS levels, and the increased MMP and AMPKα phosphorylation levels. In the metformin group, the Dnmt1, Dnmt3a, and Hdac2 genes have significantly upregulated compared to the DHEA group. However, metformin was found to have no effect on the expression level of Hdac1. In this regard, significant decrease and increase were observed in both the acetylated H4K16 and methylated H3K9 within MII oocytes in the DHEA + Metformin group compared with the DHEA group. Our results show that metformin could enhance the developmental competence of PCOS oocytes via reducing OS and ECs.

Application of Empirical Mode Decomposition for Decoding Perception of Faces Using Magnetoencephalography

Neural decoding is useful to explore the timing and source location in which the brain encodes information. Higher classification accuracy means that an analysis is more likely to succeed in extracting useful information from noises. In this paper, we present the application of a nonlinear, nonstationary signal decomposition technique—the empirical mode decomposition (EMD), on MEG data. We discuss the fundamental concepts and importance of nonlinear methods when it comes to analyzing brainwave signals and demonstrate the procedure on a set of open-source MEG facial recognition task dataset. The improved clarity of data allowed further decoding analysis to capture distinguishing features between conditions that were formerly over-looked in the existing literature, while raising interesting questions concerning hemispheric dominance to the encoding process of facial and identity information.

Revealing the Dynamic Nature of Amplitude Modulated Neural Entrainment With Holo-Hilbert Spectral Analysis

Patterns in external sensory stimuli can rapidly entrain neuronally generated oscillations observed in electrophysiological data. Here, we manipulated the temporal dynamics of visual stimuli with cross-frequency coupling (CFC) characteristics to generate steady-state visual evoked potentials (SSVEPs). Although CFC plays a pivotal role in neural communication, some cases reporting CFC may be false positives due to non-sinusoidal oscillations that can generate artificially inflated coupling values. Additionally, temporal characteristics of dynamic and non-linear neural oscillations cannot be fully derived with conventional Fourier-based analyses mainly due to trade off of temporal resolution for frequency precision. In an attempt to resolve these limitations of linear analytical methods, Holo-Hilbert Spectral Analysis (HHSA) was investigated as a potential approach for examination of non-linear and non-stationary CFC dynamics in this study. Results from both simulation and SSVEPs demonstrated that temporal dynamic and non-linear CFC features can be revealed with HHSA. Specifically, the results of simulation showed that the HHSA is less affected by the non-sinusoidal oscillation and showed possible cross frequency interactions embedded in the simulation without any a priori assumptions. In the SSVEPs, we found that the time-varying cross-frequency interaction and the bidirectional coupling between delta and alpha/beta bands can be observed using HHSA, confirming dynamic physiological signatures of neural entrainment related to cross-frequency coupling. These findings not only validate the efficacy of the HHSA in revealing the natural characteristics of signals, but also shed new light on further applications in analysis of brain electrophysiological data with the aim of understanding the functional roles of neuronal oscillation in various cognitive functions.

Frontoparietal Beta Amplitude Modulation and its Interareal Cross-frequency Coupling in Visual Working Memory

Visual working memory (VWM) relies on sustained neural activities that code information via various oscillatory frequencies. Previous studies, however, have emphasized time-frequency power changes, while overlooking the possibility that rhythmic amplitude variations can also code frequency-specific VWM information in a completely different dimension. Here, we employed the recently-developed Holo-Hilbert spectral analysis to characterize such nonlinear amplitude modulation(s) (AM) underlying VWM in the frontoparietal systems. We found that the strength of AM in mid-frontal beta and gamma oscillations during late VWM maintenance and VWM retrieval correlated with people's VWM performance. When behavioral performance was altered with transcranial electric stimulation, AM power changes during late VWM maintenance in beta, but not gamma, tracked participants' VWM variations. This beta AM likely codes information by varying its amplitude in theta period for long-range propagation, as our connectivity analysis revealed that interareal theta-beta couplings-bidirectional between mid-frontal and right-parietal during VWM maintenance and unidirectional from right-parietal to left-middle-occipital during late VWM maintenance and retrieval-underpins VWM performance and individual differences.

A Hilbert-based method for processing respiratory timeseries

In this technical note, we introduce a new method for estimating changes in respiratory volume per unit time (RVT) from respiratory bellows recordings. By using techniques from the electrophysiological literature, in particular the Hilbert transform, we show how we can better characterise breathing rhythms, with the goal of improving physiological noise correction in functional magnetic resonance imaging (fMRI). Specifically, our approach leads to a representation with higher time resolution and better captures atypical breathing events than current peak-based RVT estimators. Finally, we demonstrate that this leads to an increase in the amount of respiration-related variance removed from fMRI data when used as part of a typical preprocessing pipeline. Our implementation is publicly available as part of the PhysIO package, which is distributed as part of the open-source TAPAS toolbox (https://translationalneuromodeling.org/tapas).

Waveform changes with the evolution of beta bursts in the human subthalamic nucleus

OBJECTIVE

Phasic bursts of beta band synchronisation have been linked to motor impairment in Parkinson's disease (PD). However, little is known about what terminates bursts.

METHODS

We used the Hilbert-Huang transform to investigate beta bursts in the local field potential recorded from the subthalamic nucleus in nine patients with PD on and off levodopa.

RESULTS

The sharpness of the beta waveform extrema fell as burst amplitude dropped. Conversely, an index of phase slips between waveform extrema, and the power of concurrent theta activity increased as burst amplitude fell. Theta activity was also increased on levodopa when beta bursts were attenuated. These phenomena were associated with reduction in coupling between beta phase and high gamma activity amplitude. We discuss how these findings may suggest that beta burst termination is associated with relative desynchronization of the beta drive, increase in competing theta activity and increased phase slips in the beta activity.

CONCLUSIONS

We characterise the dynamical nature of beta bursts, thereby permitting inferences about underlying activities and, in particular, about why bursts terminate.

SIGNIFICANCE

Understanding the dynamical nature of beta bursts may help point to interventions that can cause their termination and potentially treat motor impairment in PD.