The idea of combining Artificial Intelligence research and Computational Modeling with Neuroscience and Cognitive Sciences has recently gained popularity leading to a novel research philosophy for which the term “Cognitive Computational Neuroscience” has been coined. For example, this interdisciplinary approach has already gained attention as regards the human visual system, as receptive fields in the retina served as a template for convolutional neural networks, leading to a breakthrough in image classification tasks and computer vision. In this context, deep neural networks are used as models of the visual system to predict and generate new hypotheses, tested with brain and behavioral data.
Regarding the Auditory System, the same interdisciplinary approach and close cooperation with computational sciences, would open up many interesting opportunities for Auditory Neuroscience. It is already common knowledge that there is the urge to develop novel methods to evaluate high-dimensional neuronal data, as well as new models to predict for further experimental paradigms to progress in the field. Thus, we want to further establish and strengthen the connection between AI and auditory neuroscience, in particular, in the context of auditory phantom perception such as subjective tinnitus, for instance. There is not yet sufficient understanding of stimulus processing along the auditory pathway, and no clear mechanistic explanation for phenomena such as tinnitus.
The aim of this Research Topic is to develop new computational models and tools to further advance in explaining the neural computations along the auditory pathway, in both healthy and impaired auditory systems.
Special emphasis is put on the discussion regarding the right level of description of auditory phantom perceptions, as there exist fine-grained computational models for molecular mechanisms and coarse-grained models describing the basic computation based on Bayesian Statistics. It is important to discuss what is shared by these models and how they can combine to draw the whole picture. We also aim to find novel tools to analyze the high-dimensional data gained from high density EEG/MEG, multi-electrode recordings, and fMRI, but also behavioral paradigms, helping to understand how auditory stimuli are encoded.
The goal is to bring together machine learning (ML), artificial intelligence (AI) and computational modeling researchers, with neuroscientists, psychologists and physicians pointing at the paths towards the next major directions in the field of auditory neuroscience and tinnitus research.
This collection will serve as starting point and platform to discuss further research directions to unravel the mysteries of auditory processing, tinnitus development and chronic manifestation.
We welcome all types of articles (including opinions and perspectives) addressing the following:
• computational modelling, based on artificial deep neural networks as models for sensory pathways
• explore new methods including algorithms from AI and ML to explore neuronal and behavioral data
• novel ideas from a medical point of view, which could be implemented in computational models or in innovative evaluation tools (Thus, we explicitly welcome submissions from experimental scientists providing new ideas for innovative modeling and evaluation approaches.)
The idea of combining Artificial Intelligence research and Computational Modeling with Neuroscience and Cognitive Sciences has recently gained popularity leading to a novel research philosophy for which the term “Cognitive Computational Neuroscience” has been coined. For example, this interdisciplinary approach has already gained attention as regards the human visual system, as receptive fields in the retina served as a template for convolutional neural networks, leading to a breakthrough in image classification tasks and computer vision. In this context, deep neural networks are used as models of the visual system to predict and generate new hypotheses, tested with brain and behavioral data.
Regarding the Auditory System, the same interdisciplinary approach and close cooperation with computational sciences, would open up many interesting opportunities for Auditory Neuroscience. It is already common knowledge that there is the urge to develop novel methods to evaluate high-dimensional neuronal data, as well as new models to predict for further experimental paradigms to progress in the field. Thus, we want to further establish and strengthen the connection between AI and auditory neuroscience, in particular, in the context of auditory phantom perception such as subjective tinnitus, for instance. There is not yet sufficient understanding of stimulus processing along the auditory pathway, and no clear mechanistic explanation for phenomena such as tinnitus.
The aim of this Research Topic is to develop new computational models and tools to further advance in explaining the neural computations along the auditory pathway, in both healthy and impaired auditory systems.
Special emphasis is put on the discussion regarding the right level of description of auditory phantom perceptions, as there exist fine-grained computational models for molecular mechanisms and coarse-grained models describing the basic computation based on Bayesian Statistics. It is important to discuss what is shared by these models and how they can combine to draw the whole picture. We also aim to find novel tools to analyze the high-dimensional data gained from high density EEG/MEG, multi-electrode recordings, and fMRI, but also behavioral paradigms, helping to understand how auditory stimuli are encoded.
The goal is to bring together machine learning (ML), artificial intelligence (AI) and computational modeling researchers, with neuroscientists, psychologists and physicians pointing at the paths towards the next major directions in the field of auditory neuroscience and tinnitus research.
This collection will serve as starting point and platform to discuss further research directions to unravel the mysteries of auditory processing, tinnitus development and chronic manifestation.
We welcome all types of articles (including opinions and perspectives) addressing the following:
• computational modelling, based on artificial deep neural networks as models for sensory pathways
• explore new methods including algorithms from AI and ML to explore neuronal and behavioral data
• novel ideas from a medical point of view, which could be implemented in computational models or in innovative evaluation tools (Thus, we explicitly welcome submissions from experimental scientists providing new ideas for innovative modeling and evaluation approaches.)
