Highlights • Data augmentation (DA) is increasingly used with deep learning (DL) on EEG • It enhances decoding accuracy left unexplained by 29% on average on the datasets we review • We analyze which specific DA techniques appear to work best for which EEG tasks • We tested various DA techniques on an open motor-imagery task and compared the accuracy gains to demonstrate the usefulness of DA for DL-based EEG analysis • We propose guidelines for reporting parameters for different DA techniques Abstract-Background Data augmentation (DA) has recently been demonstrated to achieve considerable performance gains for deep learning (DL)-increased accuracy and stability and reduced overfitting. Some electroencephalography (EEG) tasks suffer from low samples-to-features ratio, severely reducing DL effectiveness. DA with DL thus holds transformative promise for EEG processing, possibly like DL revolutionized computer vision, etc.-New method We review trends and approaches to DA for DL in EEG to address: Which DA approaches exist and are common for which EEG tasks? What input features are used? And, what kind of accuracy gain can be expected?-Results DA for DL on EEG begun 5 years ago and is steadily used more. We grouped DA techniques (noise addition, generative adversarial networks, sliding windows, sampling, Fourier transform, recombination of segmentation, and others) and EEG tasks (into seizure detection, sleep stages, motor imagery, mental workload, emotion recognition, motor tasks, and visual tasks). DA efficacy across techniques varied considerably. Noise addition and sliding windows provided the highest accuracy boost; mental workload most benefitted from DA. Sliding window, noise addition, and sampling methods most common for seizure detection, mental workload, and sleep stages, respectively.-Comparing with existing methods Percent of decoding accuracy explained by DA beyond unaugmented accuracy varied between 8% for recombination of segmentation and 36% for noise addition and from 14% for motor imagery to 56% for mental workload-29% on average.-Conclusions DA increasingly used and considerably improved DL decoding accuracy on EEG. Additional publications-if adhering to our reporting guidelines-will facilitate more detailed analysis.
The hippocampus is critical for episodic memory, and synaptic changes induced by long-term potentiation (LTP) are thought to underlie memory formation. In rodents, hippocampal LTP may be induced through electrical stimulation of the perforant path. To test whether similar techniques could improve episodic memory in humans, we implemented a microstimulation technique that allowed delivery of low-current electrical stimulation via 100 μm-diameter microelectrodes. As thirteen neurosurgical patients performed a person recognition task, microstimulation was applied in a theta-burst pattern, shown to optimally induce LTP. Microstimulation in the right entorhinal area during learning significantly improved subsequent memory specificity for novel portraits; participants were able both to recognize previously-viewed photos and reject similar lures. These results suggest that microstimulation with physiologic level currents—a radical departure from commonly used deep brain stimulation protocols—is sufficient to modulate human behavior and provides an avenue for refined interrogation of the circuits involved in human memory.
Background: Rapid, preoperative identification of patients with the highest risk for medical complications is necessary to ensure that limited infrastructure and human resources are directed towards those most likely to benefit. Existing risk scores either lack specificity at the patient level or utilise the American Society of Anesthesiologists (ASA) physical status classification, which requires a clinician to review the chart. Methods: We report on the use of machine learning algorithms, specifically random forests, to create a fully automated score that predicts postoperative in-hospital mortality based solely on structured data available at the time of surgery. Electronic health record data from 53 097 surgical patients (2.01% mortality rate) who underwent general anaesthesia between April 1, 2013 and December 10, 2018 in a large US academic medical centre were used to extract 58 preoperative features. Results: Using a random forest classifier we found that automatically obtained preoperative features (area under the curve [AUC] of 0.932, 95% confidence interval [CI] 0.910e0.951) outperforms Preoperative Score to Predict Postoperative Mortality (POSPOM) scores (AUC of 0.660, 95% CI 0.598e0.722), Charlson comorbidity scores (AUC of 0.742, 95% CI
The readiness potential (RP)—a key ERP correlate of upcoming action—is known to precede subjects' reports of their decision to move. Some view this as evidence against a causal role for consciousness in human decision-making and thus against free-will. But previous work focused on arbitrary decisions—purposeless, unreasoned, and without consequences. It remains unknown to what degree the RP generalizes to deliberate, more ecological decisions. We directly compared deliberate and arbitrary decision-making during a $1000-donation task to non-profit organizations. While we found the expected RPs for arbitrary decisions, they were strikingly absent for deliberate ones. Our results and drift-diffusion model are congruent with the RP representing accumulation of noisy, random fluctuations that drive arbitrary—but not deliberate—decisions. They further point to different neural mechanisms underlying deliberate and arbitrary decisions, challenging the generalizability of studies that argue for no causal role for consciousness in decision-making to real-life decisions.Editorial note: This article has been through an editorial process in which the authors decide how to respond to the issues raised during peer review. The Reviewing Editor's assessment is that all the issues have been addressed (see decision letter).
It has long been acknowledged that planar hand drawing movements conform to a relationship between movement speed and shape, such that movement speed is inversely proportional to the curvature to the power of one-third. Previous literature has detailed potential explanations for the power law's existence as well as systematic deviations from it. However, the case of speed-shape relations for three-dimensional (3D) drawing movements has remained largely unstudied. In this paper we first derive a generalization of the planar power law to 3D movements, which is based on the principle that this power law implies motion at constant equi-affine speed. This generalization results in a 3D power law where speed is inversely related to the one-third power of the curvature multiplied by the one-sixth power of the torsion. Next, we present data from human 3D scribbling movements, and compare the obtained speed-shape relation to that predicted by the 3D power law. Our results indicate that the introduction of the torsion term into the 3D power law accounts for significantly more of the variance in speed-shape relations of the movement data and that the obtained exponents are very close to the predicted values.
Maoz U, Berthoz A, Flash T. Complex unconstrained threedimensional hand movement and constant equi-affine speed. J Neurophysiol 101: 1002-1015, 2009. First published December 10, 2008 doi:10.1152/jn.90702.2008. One long-established simplifying principle behind the large repertoire and high versatility of human hand movements is the two-thirds power law-an empirical law stating a relationship between local geometry and kinematics of human hand trajectories during planar curved movements. It was further generalized not only to various types of human movements, but also to motion perception and prediction, although it was unsuccessful in explaining unconstrained three-dimensional (3D) movements. Recently, movement obeying the power law was proved to be equivalent to moving with constant planar equi-affine speed. Generalizing such motion to 3D space-i.e., to movement at constant spatial equi-affine speed-predicts the emergence of a new power law, whose utility for describing spatial scribbling movements we have previously demonstrated. In this empirical investigation of the new power law, subjects repetitively traced six different 3D geometrical shapes with their hand. We show that the 3D power law explains the data consistently better than both the two-thirds power law and an additional power law that was previously suggested for spatial hand movements. We also found small yet systematic modifications of the power-law's exponents across the various shapes, which further scrutiny suggested to be correlated with global geometric factors of the traced shape. Nevertheless, averaging over all subjects and shapes, the power-law exponents are generally in accordance with constant spatial equi-affine speed. Taken together, our findings provide evidence for the potential role of non-Euclidean geometry in motion planning and control. Moreover, these results seem to imply a relationship between geometry and kinematics that is more complex than the simple local one stipulated by the two-thirds power law and similar models.
Nutrition Examination Survey (NHANES) datasets using deep learning and machine learning algorithms. Deep-learning achieved a high performance for identifying depression on the NHANES datasets of both the United States and South Korea. Trained deep-learning and machine learning algorithms are useful for estimating the prevalence of depression.
Rational, value-based decision-making mandates selecting the option with highest subjective expected value after appropriate deliberation. We examined activity in the dorsolateral prefrontal cortex (DLPFC) and striatum of monkeys deciding between smaller, immediate rewards and larger, delayed ones. We previously found neurons that modulated their activity in this task according to the animal's choice, while it deliberated (choice neurons). Here we found neurons whose spiking activities were predictive of the spatial location of the selected target (spatial-bias neurons) or the size of the chosen reward (reward-bias neurons) before the onset of the cue presenting the decision-alternatives, and thus before rational deliberation could begin. Their predictive power increased as the values the animals associated with the two decision alternatives became more similar. The ventral striatum (VS) preferentially contained spatial-bias neurons; the caudate nucleus (CD) preferentially contained choice neurons. In contrast, the DLPFC contained significant numbers of all three neuron types, but choice neurons were not preferentially also bias neurons of either kind there, nor were spatial-bias neurons preferentially also choice neurons, and vice versa. We suggest a simple winner-take-all (WTA) circuit model to account for the dissociation of choice and bias neurons. The model reproduced our results and made additional predictions that were borne out empirically. Our data are compatible with the hypothesis that the DLPFC and striatum harbor dissociated neural populations that represent choices and predeliberation biases that are combined after cue onset; the bias neurons have a weaker effect on the ultimate decision than the choice neurons, so their influence is progressively apparent for trials where the values associated with the decision alternatives are increasingly similar.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.