Domain Adaptive Algorithm Based on Multi-Manifold Embedded Distributed Alignment for Brain-Computer Interfaces.

The use of transfer learning in brain-computer interfaces (BCIs) has potential applications. As electroencephalogram (EEG) signals vary among different paradigms and subjects, existing EEG transfer learning algorithms mainly focus on the alignment of the original space. They may not discover hidden details owing to the low-dimensional structure of EEG. To effectively transfer data from a source to target domain, a multi-manifold embedding domain adaptive algorithm is proposed for BCI. First, we aligned the EEG covariance matrix in the Riemannian manifold and extracted the characteristics of each source domain in the tangent space to reflect the differences between different source domains. Subsequently, we mapped the extracted characteristics to the Grassmann manifold to obtain a common feature representation. In domain adaptation, the geometric and statistical attributes of EEG data were considered simultaneously, and the target domain divergence matrix was updated with pseudo-labels to maximize the inter-class distance and minimize the intra-class distance. Datasets generated via BCIs were used to verify the effectiveness of the algorithm. Under two experimental paradigms, namely single-source to single-target and multi-source to single-target, the average accuracy of the algorithm on three datasets was 73.31% and 81.02%, respectively, which is more than that of several state-of-the-art EEG cross-domain classification approaches. Our multi-manifold embedded domain adaptive method achieved satisfactory results on EEG transfer learning. The method can achieve effective EEG classification without a same subject's training set.

Differences in left ventricular measurements: Attenuation versus contour based methods.

BACKGROUND: Coronary computed tomography angiography (CCTA) left ventricle (LV) volumes have prognostic value. LV measurements however can differ depending on post-processing software. Two common methods are the contour (CON) or attenuation (ATT) based methods. This study aims to determine differences in LV volume measurements using the 2 methods. METHODS: LV mid-diastolic volumes (LVMDV) were measured using both ATT and CON from 2 vendors in 750 consecutive patients undergoing CCTA. 500 were measured in a derivation cohort to establish a linear regression equation that would correct for any detected differences between the two methods. The equation was then assessed in 250 cases in the validation cohort. Comparisons were made between intra-vendor LVMDVCON and LVMDVATT as well as inter-vendor LVMDVATT. RESULTS: In the derivation cohort, the correlation between the two methods and vendors were very good (0.98 and 0.97 respectively). LVMDVCON was 20.4 ±â€¯7.4% greater than LVMDVATT. LVMDVATT was 9.2 ±â€¯6.6% greater with one vendor compared to the other. Validation cohort corrected LVMDVATT was not statistically different to measured LVMDVATT (p = 0.45). CONCLUSION: A systematic difference was found between ATT and CON measuring methods. Using a derived linear regression equation, we were able to correct for differences in measurement techniques. The method of LVMDV measurement requires careful consideration when establishing reference values and extrapolating published study results.