Biology subjects

Musslick, S.

Publications and source records attributed to Musslick, S..

5 recordsLinked to original sources

bioRxiv2025

EEG-GAN: A Generative EEG Augmentation Toolkit for Enhancing Neural Classification

Electroencephalography (EEG) is a widely applied method for decoding neural activity, offering insights into cognitive function and driving advancements in neurotechnology. However, decoding EEG data remains challenging, as classification algorithms typically require large datasets that are expensive and time-consuming to collect. Recent advances in generative artificial intelligence have enabled the creation of realistic synthetic EEG data, yet no method has consistently demonstrated that such synthetic data can lead to improvements in EEG decodability across diverse datasets. Here, we introduce EEG-GAN, an open-source generative adversarial network (GAN) designed to augment EEG data. In the most comprehensive evaluation study to date, we assessed its capacity to generate realistic EEG samples and enhance classification performance across four datasets, five classifiers, and seven sample sizes, while benchmarking it against six established augmentation techniques. We found that EEG-GAN, when trained to generate raw single-trial EEG signals, produced signals that reproduce grand-averaged waveforms and time-frequency patterns of the original data. Furthermore, training classifiers on additional synthetic data improved their ability to decode held-out empirical data. EEG-GAN achieved up to a 16% improvement in decoding accuracy, with enhancements consistent across datasets but varying among classifiers. Data augmentations were particularly effective for smaller sample sizes (30 and below), significantly improving 70% of these classification analyses and only significantly impairing 4% of analyses. Moreover, EEG-GAN significantly outperformed all benchmark techniques in 69% of the comparisons across datasets, classifiers, and sample sizes and was only significantly outperformed in 3% of comparisons. These findings establish EEG-GAN as a robust toolkit for generating realistic EEG data, which can effectively reduce the costs associated with real-world EEG data collection for neural decoding tasks.

neuroscience↗

bioRxiv2023

An Information-Theoretic Approach to Reward Rate Optimization in the Tradeoff Between Controlled and Automatic Processing in Neural Network Architectures

This article introduces a quantitative approach to modeling the cost of control in a neural network architecture when it is required to execute one or more simultaneous tasks, and its relationship to automaticity. We begin by formalizing two forms of cost associated with a given level of performance: an intensity cost that quantifies how much information must be added to the input to achieve the desired response for a given task, that we treat as the contribution of control ; and an interaction cost that quantifies the degree to which performance is degraded as a result of interference between processes responsible for performing two or more tasks, that we treat as inversely related to automaticity. We develop a formal expression of the relationship between these two costs, and use this to derive the optimal control policy for a desired level of performance. We use that, in turn, to quantify the tradeoff between control and automaticity, and suggest how this can be used as a normative framework for understanding how people adjudicate between the benefits of control and automaticity.

neuroscience↗

bioRxiv2023

Control adjustment costs limit goal flexibility: Empirical evidence and a theoretical account

A cornerstone of human intelligence is the ability to flexibly adjust our cognition and behavior as our goals change. For instance, achieving some goals requires efficiency, while others require caution. Different goals require us to engage different control processes, such as adjusting how attentive and cautious we are. Here, we show that performance incurs control adjustment costs when people adjust control to meet changing goals. Across four experiments, we provide evidence of these costs, and validate a dynamical systems model explaining the source of these costs. Participants performed a single cognitively demanding task under varying performance goals (e.g., being fast or accurate). We modeled control allocation to include a dynamic process of adjusting from ones current control state to a target state for a given performance goal. By incorporating inertia into this adjustment process, our model accounts for our empirical finding that people under-shoot their target control state more (i.e., exhibit larger adjustment costs) when goals switch rather than remain fixed (Study 1). Further validating our model, we show that the magnitude of this cost is increased when: distances between target states are larger (Study 2), there is less time to adjust to the new goal (Study 3), and goal switches are more frequent (Study 4). Our findings characterize the costs of adjusting control to meet changing goals, and show that these costs emerge directly from cognitive control dynamics. In so doing, they shed new light on the sources of and constraints on flexibility of goal-directed behavior.

neuroscience↗

bioRxiv2019

A Comparison of Non-human Primate and Deep Reinforcement Learning Agent Performance in a Virtual Pursuit-Avoidance Task

We compare the performance of non-human primates and deep reinforcement learning agents in a virtual pursuit-avoidance task, as part of an effort to understand the role that cognitive control plays in the deeply evolved skill of chase and escape behavior. Here we train two agents, a deep Q network and an actor-critic model, on a video game in which the player must capture a prey while avoiding a predator. A previously trained rhesus macaque performed well on this task, and in a manner that obeyed basic principles of Newtonian physics. We sought to compare the principles learned by artificial agents with those followed by the animal, as determined by the ability of one to predict the other. Our findings suggest that the agents learn primarily 1st order physics of motion, while the animal exhibited abilities consistent with the 2nd order physics of motion. We identify scenarios in which the actions taken by the animal and agents were consistent as well as ones in which they differed, including some surprising strategies exhibited by the agents. Finally, we remark on how the differences between how the agents and the macaque learn the task may affect their peak performance as well as their ability to generalize to other tasks.

neuroscience↗

bioRxiv2018

Motivation and Cognitive Control in Depression

Depression is linked to deficits in cognitive control and a host of other cognitive impairments arise as a consequence of these deficits. Despite of their important role in depression, there are no mechanistic models of cognitive control deficits in depression. In this paper we propose how these deficits can emerge from the interaction between motivational and cognitive processes. We review depression-related impairments in key components of motivation along with new cognitive neuroscience models that focus on the role of motivation in the decision-making about cognitive control allocation. Based on this review we propose a unifying framework which connects motivational and cognitive impairments in depression. This framework is rooted in computational models of cognitive control and offers a mechanistic understanding of cognitive control deficits in depression.

neuroscience↗