11. Conclusions and Perspectives
- The L-system formalism is suitable to implement the recurrent algorithms available in the neuroanatomical literature describing dendritic trees with a limited number of local parameters.
- The stochastic and statistical approach to computational neuroanatomy allows the generation of many non-identical neurons from a single set of parameters. This set of parameter thus describes a morphological class.
- L-Neuron uses experimental data available from the literature or measurable from morphological archives. It outputs virtual neurons in many formats compatible with graphical software and neuro-simulators.
- LN is being expanded to measure fundamental parameters directly from anatomical files and to assess which algorithm best fits the source data.
- A complete analysis of neurons generated with the 3 LN algorithms is underway to assess their applicability to various morphological classes.
- The influence of global parameters, such as tropism, is being evaluated, and other variations of neuroanatomical rules will be investigated.
- The first release of L-Neuron (for Windows and Unix) is planned for early 2000. The URL will be www.krasnow.gmu.edu/L-Neuron