Notes on Functional Imaging of Human Visual Recognition (Kanwisher et. al.)
Links
- https://web.mit.edu/bcs/nklab/media/pdfs/KanwisherChunMcDermottLeddenCogBrRes96.pdf
Main Takeaways
- Using fMRI to find functional selectivity in the brain works
- By attaching more specific regions to function, we understand more about the brain
- Face recognition is probably separate in computation and region
Notes
- Visual recognition as (1) structure itself and (2) relation between structures
- Without imaging, people with brain damage to specific areas (but the brain can reorganize with what it has, so results a bit confounding here)
- Reproduced the PET experiments with fMRI: line drawings contrasted with line drawings scrambled to find functional regions that selectively activate
- Scrambling might reveal effort, not success
- Used Kolmogorov-Smirnov test whether the regular vs scambled come from fundamentally different distributions. If so, then we can do further testing.
- Identify voxels that are below the threshold as most activated.
- Conclude with more work to look at visual ventral pathway (recognition)