ABSTRACT
We describe a method for the neural decoding of memory from EEG data. Using this method, a concept being recalled can be identified from an EEG trace with an average top-1 accuracy of about 78.4% (chance 4%). The method employs deep representation learning with supervised contrastive loss to map an EEG recording of brain activity to a low-dimensional space. Because representation learning is used, concepts can be identified even if they do not appear in the training data set. However, reference EEG data must exist for each such concept. We also show an application of the method to the problem of information retrieval. In neural information retrieval, EEG data is captured while a user recalls the contents of a document, and a list of links to predicted documents is produced.
ABSTRACT
Background Stiffness of the proximal aorta may play a critical role in adverse left ventricular (LV)-vascular interactions and associated LV diastolic dysfunction. In a community-based sample, we sought to determine the association between proximal aortic stiffness measured by cardiovascular magnetic resonance (CMR) and several clinical measures of LV diastolic mechanics. Methods and Results Framingham Heart Study Offspring adults (n=1502 participants, mean 67±9 years, 54% women) with available 1.5T CMR and transthoracic echocardiographic measures were included. Measures included proximal descending aortic strain and aortic arch pulse wave velocity by CMR (2002-2006) and diastolic function (mitral Doppler E and A wave velocity, E wave area, and LV tissue Doppler e' velocity) by echocardiography (2005-2008). Multivariable linear regression analysis was used to relate CMR aortic stiffness measures to measures of echocardiographic LV diastolic function. All continuous variables were standardized. In multivariable-adjusted regression analyses, aortic strain was inversely associated with E wave deceleration time (estimated ß=-0.10±0.032, P=0.001), whereas aortic arch pulse wave velocity was inversely associated with E/A ratio (estimated ß=-0.094±0.027, P=0.0006), E wave area (estimated ß=-0.070±0.027, P=0.010), and e' (estimated ß=-0.061±0.027, P=0.022), all indicating associations of higher aortic stiffness by CMR with less favorable LV diastolic function. Compared with men, women had a larger inverse relationship between pulse wave velocity and E/A ratio (interaction ß=-0.085±0.031, P=0.0064). There was no significant effect modification by age or a U-shaped (quadratic) relation between aortic stiffness and LV diastolic function measures. Conclusions Higher proximal aortic stiffness is associated with less favorable LV diastolic function. Future studies may clarify temporal relations of aortic stiffness with varying patterns and progression of LV diastolic dysfunction.
Subject(s)
Vascular Stiffness , Ventricular Dysfunction, Left , Male , Humans , Adult , Female , Pulse Wave Analysis , Ventricular Function, Left , Ventricular Dysfunction, Left/diagnostic imaging , Ventricular Dysfunction, Left/epidemiology , DiastoleABSTRACT
[Figure: see text].
