Correcting False Memories: The Effect of Mnemonic Generalization on Original Memory Traces

Correcting False Memories: The Effect of Mnemonic Generalization on Original Memory Traces

ABSTRACT
False memories are a common occurrence but the impact of misremembering on the original memory trace is ill-described. While the original memory may be rewritten, it is also possible for a second false memory to exist concurrently with the original, and if a false memory exists concurrently then recovery of the original information should be possible. This study investigates first, whether false recognition overwrites the original memory representation using a mnemonic discrimination task, and second, which neural processes are involved in recovering the original memory following a false memory. Thirty-five healthy, young adults performed multiple recognition memory tests, where the design of the experiment induced participants to make memory errors in the first recognition memory test and then allowed us to determine whether the memory error would be corrected in the second test session. FMRI signal associated with the encoding and retrieval processes during the experiment were investigated in order to determine the important regions for false memory correction. We found that false memories do not overwrite the original trace in all instances, as recovery of the original information was possible. Critically, we determined that recovery of the original information was dependent on higher-order processes during the formation of the false memory during the first test, and not on processing at the time of encoding or the second test episode.

MRI METHODS
Functional and structural images were acquired on a Siemens TIM Trio 3T MRI scanner utilizing the 32-channel head coil. Participants contributed a T1-weighted scan, a high-resolution T2-weighted scan, and a series of functional T2*-weighted scans as described below. Standard-resolution structural images were acquired using a T1-weighted magnetization-prepared rapid acquisition with gradient echo (MP-RAGE) sequence with the following parameters: TR = 1900 ms, TE = 2.26 ms, flip angle = 9°, FoV = 218 × 250, voxel size = 0.97 × 0.97 × 1 mm, slices = 176 interleaved. High-resolution structural images were acquired using a T2-weighted sequence with the following parameters: TR = 8020 ms, TE = 80 ms, flip angle = 150°, FoV = 150 × 150 mm, voxel size = 0.4 × 0.4 × 2 mm, slices = 30 interleaved. The T2 data were acquired perpendicular to the long axis of the hippocampus. High-resolution multi-band echo-planar images were acquired with a T2*-weighted pulse sequence with the following parameters: Multi-band factor = 8, TR = 875 ms, TE = 43.6 ms, flip angle = 55°, FoV = 180 × 180 mm, voxel size = 1.8 mm3, slices = 72 interleaved. This sequence was aligned parallel with the long axis of the hippocampus, and the first 11 TRs were discarded in order to allow for T1 equilibration.