Tianlu Wang

Neural communication across different spatial and temporal scales is a topic of great interest in clinical and basic science. Phase‐amplitude coupling (PAC) has attracted particular interest due to its functional role in a wide range of cognitive and motor functions. Here, we introduce a novel measure termed the direct modulation index (dMI). Based on the classical modulation index, dMI provides an estimate of PAC that is (1) bound to an absolute interval between 0 and +1, (2) resistant against noise, and (3) reliable even for small amounts of data. To highlight the properties of this newly‐proposed measure, we evaluated dMI by comparing it to the classical modulation index, mean vector length, and phase‐locking value using simulated data. We ascertained that dMI provides a more accurate estimate of PAC than the existing methods and that is resilient to varying noise levels and signal lengths. As such, dMI permits a reliable investigation of PAC, which may reveal insights crucial to our understanding of functional brain architecture in key contexts such as behaviour and cognition. A Python toolbox that implements dMI and other measures of PAC is freely available at https://github.com/neurophysiological-analysis/FiNN.

Communication between neural structures is a topic of much clinical and scientific interest and has been linked to a variety of behavioural, cognitive, and psychiatric measures. Here, we introduce a novel effective connectivity measure, termed the directional absolute coherence (DAC). Combining aspects of magnitude squared coherence, imaginary coherence, and phase slope index, DAC provides an estimate of connectivity that is resistant to volume conduction, encapsulates the directionality of neural communication, and is bound to the interval of –1 and 1. To highlight the properties of this newly proposed method, we compare DAC to a number of established connectivity methods using data recorded from the subthalamic nucleus of patients with Parkinson’s disease with deep brain stimulation electrodes. By applying a combination of real and simulated data, we demonstrate that DAC provides a reliable estimate of the magnitude and direction of connectivity, independent of the phase difference between brain signals. As such, DAC facilitates a reliable investigation of inter-regional neural communication, rendering it a valuable tool for gaining a deeper understanding of the functional architecture of the brain and its relationship to behaviour and cognition. A Python implementation of DAC is freely available at https://github.com/neurophysiological-analysis/FiNN . Highlights: - Volume conduction and limited interpretability affect many connectivity methods. - DAC is a combined approach to overcome these pitfalls. - DAC augments information content and interpretability in comparison to other methods. - DAC allows for reliable estimation of effective connectivity.

Neural communication across different spatial and temporal scales is a topic of great interest in clinical and basic science. Phase-amplitude coupling (PAC) has attracted particular interest due to its functional role in a wide range of cognitive and motor functions. Here, we introduce a novel measure termed the direct modulation index (dMI). Based on the classical modulation index, dMI provides an estimate of PAC that is bound to an absolute interval between 0 and +1, resistant against noise, and reliable even for small amounts of data. To highlight the properties of this newly-proposed measure, we evaluated dMI by comparing it to the classical modulation index, mean vector length, and phase-locking value using simulated data. We ascertained that dMI provides a more accurate estimate of PAC than the existing methods and that is resilient to varying noise levels and signal lengths. As such, dMI permits a reliable investigation of PAC, which may reveal insights crucial to our understanding of functional brain architecture in key contexts such as behaviour and cognition. A Python toolbox that implements dMI and other measures of PAC is freely available at https://github.com/neurophysiological-analysis/FiNN. Highlights - Neural coupling measures are sensitive to higher harmonics of the target oscillation. - dMI achieves frequency-specificity by sine-fitting the phase-amplitude histogram. - Increased robustness to noise and signal duration in comparison to other measures. - dMI allows for reliable estimation of phase-amplitude coupling.

Objective: Visual attention helps us to selectively process relevant information and is crucial in our everyday interactions with the environment. Not surprisingly, it is one of the cognitive domains that is most frequently affected by acquired brain injury. Reliable assessment of attention deficits is pivotal to neuropsychological examination and helps to optimize individual rehabilitation plans. Compared with conventional pen-and-paper tests, computerized tasks borrowed from the field of experimental psychology bring many benefits, but lab-based experimental setups cannot be easily incorporated in clinical practice. Light-weight and portable mobile tablet devices may facilitate the translation of computerized tasks to clinical settings. One such task is based on the Theory of Visual Attention (TVA), a mathematical model of visual attention. TVA-based paradigms have been widely used to investigate several aspects of visual attention in both fundamental and clinical research, and include measures for general processing capacity as well as stimulus-specific attentional parameters. Methods: This article discusses the benefits of TVA-based assessments compared with frequently used neuropsychological tests of visual attention, and examines the reliability of a tablet-based TVA-based assessment in 59 neurologically healthy participants. Results: Pearson’s correlations indicate that the tablet-based TVA assessment and the conventional lab-based TVA assessment have a comparable parallel-form (range: .67–.93), test–retest (range: .61–.78), and internal reliability (range: .56–.97). Conclusion: Our results suggest that tablet-based TVA assessment may be a promising tool to acquire clinical measures of visual attention at low cost at the bedside of the patient.

NLP models are shown to suffer from robustness issues, i.e., a model's prediction can be easily changed under small perturbations to the input. In this work, we present a Controlled Adversarial Text Generation (CAT-Gen) model that, given an input text, generates adversarial texts through controllable attributes that are known to be invariant to task labels. For example, in order to attack a model for sentiment classification over product reviews, we can use the product categories as the controllable attribute which would not change the sentiment of the reviews. Experiments on real-world NLP datasets demonstrate that our method can generate more diverse and fluent adversarial texts, compared to many existing adversarial text generation approaches. We further use our generated adversarial examples to improve models through adversarial training, and we demonstrate that our generated attacks are more robust against model re-training and different model architectures.

We propose an inference procedure for deep convolutional neural networks (CNNs) when partial evidence is available. Our method consists of a general feedback-based propagation approach (feedback-prop) that boosts the prediction accuracy for an arbitrary set of unknown target labels when the values for a non-overlapping arbitrary set of target labels are known. We show that existing models trained in a multi-label or multi-task setting can readily take advantage of feedback-prop without any retraining or fine-tuning. Our feedback-prop inference procedure is general, simple, reliable, and works on different challenging visual recognition tasks. We present two variants of feedback-prop based on layer-wise and residual iterative updates. We experiment using several multi-task models and show that feedback-prop is effective in all of them. Our results unveil a previously unreported but interesting dynamic property of deep CNNs. We also present an associated technical approach that takes advantage of this property for inference under partial evidence in general visual recognition tasks.