In this lecture, we will analyze how different biomathematical models of the evolution of intracerebral gliomas and their interaction with brain function can - or may in the near future - help the neurosurgeon at different stages of patient management.
In the pre-operative phase, the proliferation-diffusion tumor growth model, fed by consecutive MRI scans of a patient, can be used to determine individual growth kinetics, a fundamental element in surgical decision-making. On the other hand, the functional resectability atlas - constructed from postoperative residuals of patient series operated on in reference centers, and designed to objectively estimate whether a patient's tumor is operable or not - provides a privileged tool for homogenizing management strategies between different institutions.
During the operation, the surgeon uses neuronavigation as an intracerebral GPS. However, the "brain shift" generated by the collapse of the brain as the operation proceeds, considerably limits the precision of this tool (up to 1 cm offset). Real-time correction algorithms, taking into account the biomechanical properties of brain tissue, would provide the surgeon with undeniable assistance. In addition, all tumor and functional mappings from multi-sequence MRI can now be integrated into neuronavigation. Among these, white matter bundle tractographies, resulting from the application of tracking algorithms based on diffusion tensor sequences, provide invaluable information on the neural networks to be preserved. As we shall see, however, the anatomical variability of clusters obtained by different tracking algorithms considerably limits the value of this imaging technique, especially as functional imaging of these clusters is not currently available. This is why awake surgery has become the standard for operating on gliomas in functional areas. This methodology offers a unique opportunity to study the principles of brain function. In particular, axon-cortical evoked potentials, recorded in response to single pulses applied to the bundles, provide a powerful means of identifying functional neuronal networks.
