• Digital Humanities and Cultural Heritage
• EBRAINS close

AbstractMore than 100 years of research have passed by and still the human electroencephalogram (EEG) remains a puzzle to be solved. Starting from his studies on plethysmography until his theories on brain thermodynamics, Hans Berger was able to refine his method of recording cortical signs with the apparatus at his disposal in an ordinary neuropsychiatric yard towards an early account of human EEG. This review is an appraisal of his contribution to the field of modern neurophysiology.

Abstract Retracing human cognitive origins started out at the systems level with the top-down interpretation of archaeological records spanning from man-made artifacts to endocasts of ancient skulls. With emerging evolutionary genetics and organoid technologies, it is now possible to deconstruct evolutionary processes on a molecular/cellular level from the bottom-up by functionally testing archaic alleles in experimental models. The current challenge is to complement these approaches with novel strategies that allow a holistic reconstruction of evolutionary patterns across human cognitive domains. We argue that computational neuroarcheology can provide such a critical mesoscale framework at the brain network-level, linking molecular/cellular (bottom-up) to systems (top-down) level data for the correlative archeology of the human mind.

AbstractDeep learning has become a leading subset of machine learning and has been successfully employed in diverse areas, ranging from natural language processing to medical image analysis. In medical imaging, researchers have progressively turned towards multi-center neuroimaging studies to address complex questions in neuroscience, leveraging larger sample sizes and aiming to enhance the accuracy of deep learning models. However, variations in image pixel/voxel characteristics can arise between centers due to factors including differences in magnetic resonance imaging scanners. Such variations create challenges, particularly inconsistent performance in machine learning-based approaches, often referred to as domain shift, where the trained models fail to achieve satisfactory or improved results when confronted with dissimilar test data. This study analyzes the performance of multiple disease classification tasks using multi-center MRI data obtained from three widely used scanner manufacturers (GE, Philips, and Siemens) across several deep learning-based networks. Furthermore, we investigate the efficacy of mitigating scanner vendor effects using ComBat-based harmonization techniques when applied to multi-center datasets of 3D structural MR images. Our experimental results reveal a substantial decline in classification performance when models trained on one type of scanner manufacturer are tested with data from different manufacturers. Moreover, despite applying ComBat-based harmonization, the harmonized images do not demonstrate any noticeable performance enhancement for disease classification tasks.

Summary: Recent work suggests that serial dependence, where perceptual decisions are biased toward previous stimuli, arises from the prior that sensory input is temporally correlated. However, existing studies have mostly used random stimulus sequences that do not involve such temporal consistencies. Here, we manipulated the temporal statistics of visual stimuli to examine the role of true temporal correlations in serial dependence. In two experiments, observers reproduced the orientation of the last stimulus in a sequence, while we varied temporal correlations in the stimulus features at two timescales: stimulus history within the trial and decision history across trials. We found a clear dissociation: increasing temporal correlation in the stimulus history led to adaptation-like repulsive biases, whereas increasing temporal correlation in the decision history reduced attractive biases. Thus, we suggest that temporal correlation enhances the discriminative ability of the visual system, revealing the fundamental role of the broader temporal context.

En 1810, Gall et Spurzheim posent le diagnostic suivant : la piètre connaissance du substrat cérébral des facultés mentales doit être imputée à la longue inféodation de l’anatomie par la métaphysique. Dans le présent article, nous interrogeons un tel jugement historique à partir du Discours sur l’anatomie du cerveau de Niels Steensen. La présentation des « systèmes » de localisation cérébrale que Steensen critique (celui des « Anciens », celui de Thomas Willis et celui de Descartes) nous permettra d’abord de brosser un tableau des différentes conceptions possibles du « siège de l’âme » au XVIIe siècle. Nous montrerons ensuite comment, dans ces trois cas de figure, la mise en correspondance d’une anatomie neurocérébrale lacunaire et d’une topique des grandes facultés mentales conduit à postuler que telle ou telle partie cérébrale est le siège de la mémoire, celui de l’imagination ou encore celui du sens commun. Enfin, nous montrerons que c’est au nom de la science anatomique, restreinte à ce qui peut être attesté par des témoins oculaires, que Steensen a écarté le principe même des localisations cérébrales. In 1810, Gall and Spurzheim proposed the following diagnosis: the scant of knowledge of the cerebral substratum responsible for mental faculties must be attributed to the excessive subservience of anatomy to metaphysics. In this article, we examine this historic judgement with reference to Niels Steensen’s Discourse on the Anatomy of the Brain. By presenting the cerebral localisation “systems” of which Steensen is critical (those proposed by the “ancients”, by Thomas Willis and Descartes), we will firstly provide an overview of the different conceptions of the “seat of the soul” in the seventeenth century. We will then show how, in these three cases, the correlation between a lacunar brain mapping and a spatial representation of mental faculties leads to the postulation that a particular part of the brain is the seat of memory, imagination or common sense. Finally, we will show that it was in the name of anatomical science, limited to what can be evidenced by eyewitnesses, that Steensen rejected the very principle of cerebral localisation. International audience

Over the past decade, the artificial neural networks domain has seen a considerable embracement of deep neural networks among many applications. However, deep neural networks are typically computationally complex and consume high power, hindering their applicability for resource-constrained applications, such as self-driving vehicles, drones, and robotics. Spiking neural networks, often employed to bridge the gap between machine learning and neuroscience fields, are considered a promising solution for resource-constrained applications. Since deploying spiking neural networks on traditional von-Newman architectures requires significant processing time and high power, typically, neuromorphic hardware is created to execute spiking neural networks. The objective of neuromorphic devices is to mimic the distinctive functionalities of the human brain in terms of energy efficiency, computational power, and robust learning. Furthermore, natural language processing, a machine learning technique, has been widely utilized to aid machines in comprehending human language. However, natural language processing techniques cannot also be deployed efficiently on traditional computing platforms. In this research work, we strive to enhance the natural language processing traits/abilities by harnessing and integrating the SNNs traits, as well as deploying the integrated solution on neuromorphic hardware, efficiently and effectively. To facilitate this endeavor, we propose a novel, unique, and efficient sentiment analysis model created using a large-scale SNN model on SpiNNaker neuromorphic hardware that responds to user inputs. SpiNNaker neuromorphic hardware typically can simulate large spiking neural networks in real time and consumes low power. We initially create an artificial neural networks model, and then train the model using an Internet Movie Database (IMDB) dataset. Next, the pre-trained artificial neural networks model is converted into our proposed spiking neural networks model, called a spiking sentiment analysis (SSA) model. Our SSA model using SpiNNaker, called SSA-SpiNNaker, is created in such a way to respond to user inputs with a positive or negative response. Our proposed SSA-SpiNNaker model achieves 100% accuracy and only consumes 3970 Joules of energy, while processing around 10,000 words and predicting a positive/negative review. Our experimental results and analysis demonstrate that by leveraging the parallel and distributed capabilities of SpiNNaker, our proposed SSA-SpiNNaker model achieves better performance compared to artificial neural networks models. Our investigation into existing works revealed that no similar models exist in the published literature, demonstrating the uniqueness of our proposed model. Our proposed work would offer a synergy between SNNs and NLP within the neuromorphic computing domain, in order to address many challenges in this domain, including computational complexity and power consumption. Our proposed model would not only enhance the capabilities of sentiment analysis but also contribute to the advancement of brain-inspired computing. Our proposed model could be utilized in other resource-constrained and low-power applications, such as robotics, autonomous, and smart systems.

According to contemporary computational neuroscience the mental is associated with computations implemented in the brain. We analyze in physical terms based on recent results in the foundations of statistical mechanics two well-known (independent) problems that arise for this approach: the problem of multiple-computations and the problem of multiple-realization. We show that within the computational theory of the mind the two problems are insoluble by the physics of the brain. We further show that attempts to solve the problems by the interactions of the systems implementing the computations with an environment (in or outside the brain) must introduce non-physical factors, and therefore fail on physical grounds. We also show that the problems are endemic and pertain to other forms of functional theories of the mind, most notably, causal functionalism. Finally, we propose a physicalist reductive identity theory, which is a generalization of statistical mechanics for all the special sciences, and show that only a theory of this kind can provide physical solutions to the above two problems in computational neuroscience. We conclude that functionalism in the theory of mind must be replaced with a reductive identity theory. This result has far-reaching implications with respect to the research programs in brain science.