So many examples of brain localization when in reality only %30 of the brain is accurately mapped. What has been involved in the difficulty for all these years? Shouldn't understanding regions of neuronal activity be enough to make accurate comparisons to other brain models?
Gary Covin that is a great question and actually a critical view at a long-standing controversy!
My opinion:
The accuracy and probity assessment in systematic brain mapping approaches seems to be rife with loop holes in my opinion because brain structure is plastic, and thus its function could lead to heterogeneity in sampling and outliers in analysis or processing that would make the interpretation of data quite tricky and sometime problematic or paradoxical.
Besides, it seems that the results of different mapping modalities and their often times competitive approaches to analyze and interpret data have not yet reached a confirmed reconciliation. They simply show that they cannot yet corroborate results of other cellular level approaches.
Even though ROIs and in general regions of activity in brain can be processed with a consistently homogeneous model through out its entire setting, different models can still have distinctive characteristics based on the parameters set to primary evaluate and emulate a specific kind of brain activity even in the same ROI.
Although Siamese network as an ANN employs the same weights of parameters but I think that using them in a tandem setting on disparate input vectors to compare output ones can cause such heterogeneity in results as seen in face verification among a sizable and huge sample or data.
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping