Information Processing in the Nervous System. On a Holistic Functional Organization Principle

Sammanfattning: This thesis is concerned with the nature of information processing in the nervous system. This issue has been investigated with animal experiments, using male Sprague–Dawley rats, and simulated touch generated by an artificial fingertip with neuromorphic tactile sensors. The processing of the tactile input from the artificial touch system has been analyzed in light of two opposing theories on information integration in the nervous system. The commonly accepted theory of functional localization, where distinct functions are believed to reside in localized clusters of neurons (e.g. the specific digit area of the primary somatosensory cortex); and the opposing holistic theory where function is seen as a result of the net activity of the nervous system, i.e. as an integrated whole. In addition to the experimental studies, a simulation study has been made to evaluate the validity of a selforganizational formative principle for the connectivity schemes seen in the spinal cord. Self-organization is a cornerstone for the holistic theory and hence the inclusion. The above outline has resulted in four papers. In paper I a dense interaction between the neocortical representations of individual digits was shown, questioning the discrete nature of historically popular cortical maps. In paper II this was further expanded when complex responses to tactile stimuli was found across the whole dorsal surface of the neocortex. The interdependence between these findings were implicated in paper III in which we showed that a stroke in a distant cortical area had a negative impact on the decoding performance of tactile stimuli in the primary somatosensory cortex, thus serving as support for the holistic theory of an indivisible neocortical modus operandi. Finally, in paper IV an initial argument was made for self-organization rather than detailed preprogramming as the formative rule for circuit formation. Paper IV also included the development of a novel interpretation of an artificial neural network as well as a model organism. The findings of the included papers are contextualized with a historical background looking back more than 2000 years and also with a theoretical background in neuroscience, genetics and mathematics. All in all trying to present a multifaceted but relevant discourse on the topic of information integration in the nervous system. The results of the included papers are in favour of the holistic interpretation. A holistic functional brain organization must thus be seen as plausible and with it follows consequences for how information processing can occur in the network of the nervous system and also how a functional decomposition might be possible, or impossible, to perform. One such consequence is that the neocortex should be seen as an indivisible unit of a dynamic system.