About this Research Topic
The spark of the special issue presented here stems from the work we developed together with Prof. Glyn Humphreys who passed on 14th of January 2016. Since I started working as a postdoctoral fellow in Prof. Humphreys’ lab we had discussions about computational modelling, cognition and neuropsychology. Looking on simulating neuropsychology not only by inactivating a group of neurons from the model to simulate brain lesion but also by investigating changes in neuronal properties stemmed from neurotransmitter depletion (for example Alzheimer’s, Parkinson’s and ADHD). Then the developed model can be used not only to predict behaviour but also to assist in rehabilitation. Aspects of this work are incorporated in this special issue that the proposal and run was a joined effort of late Prof. Humprheys and I.
Dr Eirini Mavritsaki
Neuropsychology has traditionally examined impairments of the brain and behaviour, comparing these impairments with normal behaviour and brain function. Computational neuroscience came to change this in early 1990s by introducing computational modelling as a new method for unveiling the secrets of both normal and diseased brains (Bullinaria & Chater, 1995; Caramazza & Coltheart, 2006; Harley, 1993; Plaut, 1996; Plaut & Shallice, 1993). In computational modelling, theories are tested by building a mathematical model of the studied behaviour. That model simulates the studied behaviour and can be assessed using both behavioural and neural data. Such models can then be lesioned to simulate the impaired brain function and to test the original theory – rather like using a microscope in biomedical studies, where the virus is taken out of the body in an environment that simulates the body function.
Since the early 1990s, there has been a rise in neuropsychological studies using computational modelling in an effort to further understand neurological impairments. For example, computational modelling has been used in studies of language and reading disorders (Foygel & Dell, 2000; Harley, 1993; Knobel & Caramazza, 2007; Plaut, 1996; Plaut & Shallice, 1993; Ralph, 2004; Schwartz & Dell, 2010; Woollams, 2014), object recognition (Humphreys & Forde, 2001; McClelland & Rogers, 2003), empathy (Decety & Jackson, 2004), learning (Nomura & Reber, 2008; Plaut, 1996), memory (Devlin, Gonnerman, Andersen, & Seidenberg, 1998; Kliegel, Altgassen, Hering, & Rose, 2011) and attention (Ludwig, Butler, Rossit, Harvey, & Gilchrist, 2009; Mavritsaki, Heinke, Allen, Deco, & Humphreys, 2011). Furthermore, the computational modelling studies in neuropsychology have moved from examining the effects of lesioning to studying changes in neurotransmitters in the brain (Huys & Dayan, 2008; Mavritsaki et al., 2011) and to comparing the traditional neuropsychological studies with fMRI studies (Huys & Dayan, 2008; Mavritsaki et al., 2011).
We conclude that computational modelling may provide an increasingly important tool for furthering neuropsychology and understanding brain impairments – but there remain many issues: how can different types of lesion be best modelled? What are the differences between damage to grey matter and to fibre tracts? What the effects of learning within damaged systems or systems where there is an imbalance in neurotransmitters? In this Research Topic we wish to bring together state-of-the-art papers on the computational modelling of neuropsychological disorders, with papers covering different levels of modelling (from spiking neurons to higher-level connectionist modelling), using imaging as well as behavioural data, and addressing a number of different disorders (attention, language, memory etc.). The papers should highlight commonalities and differences across the different approaches and provide an important focus for future research.
Dr Eirini Mavritsaki
Neuropsychology has traditionally examined impairments of the brain and behaviour, comparing these impairments with normal behaviour and brain function. Computational neuroscience came to change this in early 1990s by introducing computational modelling as a new method for unveiling the secrets of both normal and diseased brains (Bullinaria & Chater, 1995; Caramazza & Coltheart, 2006; Harley, 1993; Plaut, 1996; Plaut & Shallice, 1993). In computational modelling, theories are tested by building a mathematical model of the studied behaviour. That model simulates the studied behaviour and can be assessed using both behavioural and neural data. Such models can then be lesioned to simulate the impaired brain function and to test the original theory – rather like using a microscope in biomedical studies, where the virus is taken out of the body in an environment that simulates the body function.
Since the early 1990s, there has been a rise in neuropsychological studies using computational modelling in an effort to further understand neurological impairments. For example, computational modelling has been used in studies of language and reading disorders (Foygel & Dell, 2000; Harley, 1993; Knobel & Caramazza, 2007; Plaut, 1996; Plaut & Shallice, 1993; Ralph, 2004; Schwartz & Dell, 2010; Woollams, 2014), object recognition (Humphreys & Forde, 2001; McClelland & Rogers, 2003), empathy (Decety & Jackson, 2004), learning (Nomura & Reber, 2008; Plaut, 1996), memory (Devlin, Gonnerman, Andersen, & Seidenberg, 1998; Kliegel, Altgassen, Hering, & Rose, 2011) and attention (Ludwig, Butler, Rossit, Harvey, & Gilchrist, 2009; Mavritsaki, Heinke, Allen, Deco, & Humphreys, 2011). Furthermore, the computational modelling studies in neuropsychology have moved from examining the effects of lesioning to studying changes in neurotransmitters in the brain (Huys & Dayan, 2008; Mavritsaki et al., 2011) and to comparing the traditional neuropsychological studies with fMRI studies (Huys & Dayan, 2008; Mavritsaki et al., 2011).
We conclude that computational modelling may provide an increasingly important tool for furthering neuropsychology and understanding brain impairments – but there remain many issues: how can different types of lesion be best modelled? What are the differences between damage to grey matter and to fibre tracts? What the effects of learning within damaged systems or systems where there is an imbalance in neurotransmitters? In this Research Topic we wish to bring together state-of-the-art papers on the computational modelling of neuropsychological disorders, with papers covering different levels of modelling (from spiking neurons to higher-level connectionist modelling), using imaging as well as behavioural data, and addressing a number of different disorders (attention, language, memory etc.). The papers should highlight commonalities and differences across the different approaches and provide an important focus for future research.
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
