Using spatial filters to study regional reorganization of brain connectivity in familial early onset AD (PSEN1 E280A)

Background The most common cause of familial early onset Alzheimer Disease (AD) are the PSEN1 mutations. One of the largest family carriers of the mutation lives in Antioquia, Colombia. The study of mutation carriers provides a unique opportunity to characterize the preclinical changes and the brain reorganization strategies associated with predisposition to AD. The Electroencephalography (EEG) has an enormous potential to study the neurophysiological changes in neurodegenerative processes with promising clinical applications. During the 2012 – 2019 period different studies on a large database of high‐density EEG from PSEN1 E280A carriers were done trying to identify early changes that could be preclinical biomarkers of AD Method A sample of 83 participants, belonging to families with genetic risk for AD by PSEN1 E280A mutation was included. They were distributed in three groups; asymptomatic carriers (AC, n=27), asymptomatic non‐carriers (ANC, n=33) and a symptomatic group with Mild Cognitive Impairment (SYMP, n=23). We used a 60 channel EEG for recording the brain electrical activity. The relative power values were obtained using Matlab’s chronux toolbox. The results obtained using a workflow based on Independent Component Analysis (ICA), introduced to improve the spatial resolution given by EEG, are discussed. A two‐parameter non‐parametric T‐test was used to calculate the statistical significance of the relative power among the populations analyzed. Result The comparison of carriers of PSEN1 E280A mutation and non‐carriers, show altered brain function even in asymptomatic stage complementing previous findings using functional Magnetic Resonance Imaging or Positron Emission Tomography. Regions, like the Precuneus, can be screened across the neurodegenerative process, offering new potential biomarkers of AD. The multivariate analysis of brain sources show generalized impairment in the brain connectivity for asymptomatic subjects. The results found using the workflow were compared with alternatives approaches of analysis, based on scalp EEG or data obtained through inverse solution methods, corroborating the findings. Conclusion The neurophysiological correlates found opens the possibility of studying new biomarkers in AD based in EEG. The workflow proposed could be used in other neurodegenerative diseases or other brain mapping problems based in EEG.