Project description

What can we learn from brain damages about numerical processes in the unaffected brain? This was the central issue of this project.

Based on the triple-code model (Dehaene, 1992; Dehaene & Cohen 1995, 1997), two basic processes are differentiated in numerical cognition – the processing of numerical magnitude information on the one hand and arithmetic fact retrieval on the other. How distinct these processes actually are, and which neuro-cognitive evidence supports this distinction, is still controversially discussed.

In order to advance this discourse, static and dynamic aspects (neuro-cognitive plasticity) of numerical magnitude processing and arithmetic fact retrieval were assessed in stroke patients and healthy controls via parallelized cross-sectional and longitudinal experimental (training-) studies. The functional-anatomical foundations of both processes were investigated by means of multivariate lesion analysis and diffusion tensor imaging (DTI).

Multivariate lesion analysis allows for differential causal structure-function mapping of functional deficits in numerical (sub)processes and lesioned parts of the postulated networks underlying numerical cognition. Furthermore, potential re-organization processes within and between the neural networks of numerical magnitude and arithmetic fact retrieval can be investigated as well (i.e., whether certain cognitive (sub)processes can be taken over by other neuro-anatomical regions to compensate for functional deficits). By means of DTI possible fiber disconnections can be identified and analyzed, which might impair functionally unimpaired brain regions.

The parallelized experimental design enabled us to address identical questions regarding magnitude processing and arithmetic fact retrieval in stroke patients and healthy controls. On this occasion, analyzing different patient groups by means of multivariate lesion analysis and DTI allowed to identify causal relationships of cognitive processing and the underlying neural correlates. At the same time, the influence of pathological interferences could be excluded by examining healthy controls in parallelized experiments. Spatial-numerical associations were examined as an indicator for the processing of numerical magnitude.

In sum, the results gained from this project could extend our understanding of (un)impaired number processing, permit prognostic statements and allow for implications for diagnosis and interventions of numerical impairments.

Cooperations

• Prof. Dr. Dr. Hans-Otto Karnath, Universitätsklinikum Tübingen