Association of volatile anesthesia exposure and depth with emergence agitation and delirium in children: Prospective observational cohort study.

Background: Sevoflurane anesthesia is widely used in pediatric ambulatory surgery. However, emergency agitation (EA) and emergency delirium (ED), as major complications following sevoflurane anesthesia in children, pose risks to surgery and prognosis. Identifying the high risk of EA/ED, especially anesthesia exposure and the depth of anesthesia, may allow preemptive treatment. Methods: A total of 137 patients were prospectively enrolled in this single-center observational cohort study to assess the incidence of EA or ED. Multivariable logistic regression analyses were used to test the association between volatile anesthesia exposure and depth with EA or ED. The Richmond Agitation and Sedation Scale (RASS), Pediatric Anesthesia Emergence Delirium Scale (PAED) and Face, Legs, Activity, Cry, and Consolability (FLACC) behavioural pain scale was used to assess the severity of EA or ED severity and pain. Bispectral index (BIS) to monitor the depth of anesthesia, as well as TimeLOW-BIS/TimeANES %, EtSevo (%) and EtSevo-time AUC were included in the multivariate logistic regression model as independent variables to analyze their association with EA or ED. Results: The overall prevalence of EA and ED was 73/137 (53.3%) and 75/137 (54.7%) respectively, where 48/137 (35.0%), 19/137 (13.9%), and 6/137 (4.4%) had mild, moderate, and severe EA. When the recovery period was lengthened, the prevalence of ED and extent of FLACC decreased and finally normalized within 30 min in recovered period. Multivariable logistic regression demonstrated that intraoperative agitation [2.84 (1.08, 7.47) p = 0.034], peak FLACC [2.56 (1.70, 3.85) p < 0.001] and adverse event (respiratory complications) [0.03 (0.00, 0.29) p = 0.003] were independently associated with higher odds of EA. Taking EtSevo-time AUC ≤ 2,000 as a reference, the incidence of EA were [15.84 (2.15, 116.98) p = 0.002] times and 16.59 (2.42, 113.83) p = 0.009] times for EtSevo-time AUC 2,500-3,000 and EtSevo-time AUC > 3,000, respectively. Peak FLACC [3.46 (2.13, 5.62) p < 0.001] and intraoperative agitation [5.61 (1.99, 15.86) p = 0.001] were independently associated with higher odds of developing ED. EtSevo (%), intraoperative BIS value and the percentage of the duration of anesthesia at different depths of anesthesia (BIS ≤ 40, BIS ≤ 30, BIS ≤ 20) were not associated with EA and ED. Conclusions: For pediatrics undergoing ambulatory surgery where sevoflurane anesthesia was administered, EA was associated with surgical time, peak FLACC, respiratory complications, and "EtSevo-time AUC" with a dose-response relationship; ED was associated with peak FLACC and intraoperative agitation.

Dimensionality Reduction of SPD Data Based on Riemannian Manifold Tangent Spaces and Isometry.

Symmetric positive definite (SPD) data have become a hot topic in machine learning. Instead of a linear Euclidean space, SPD data generally lie on a nonlinear Riemannian manifold. To get over the problems caused by the high data dimensionality, dimensionality reduction (DR) is a key subject for SPD data, where bilinear transformation plays a vital role. Because linear operations are not supported in nonlinear spaces such as Riemannian manifolds, directly performing Euclidean DR methods on SPD matrices is inadequate and difficult in complex models and optimization. An SPD data DR method based on Riemannian manifold tangent spaces and global isometry (RMTSISOM-SPDDR) is proposed in this research. The main contributions are listed: (1) Any Riemannian manifold tangent space is a Hilbert space isomorphic to a Euclidean space. Particularly for SPD manifolds, tangent spaces consist of symmetric matrices, which can greatly preserve the form and attributes of original SPD data. For this reason, RMTSISOM-SPDDR transfers the bilinear transformation from manifolds to tangent spaces. (2) By log transformation, original SPD data are mapped to the tangent space at the identity matrix under the affine invariant Riemannian metric (AIRM). In this way, the geodesic distance between original data and the identity matrix is equal to the Euclidean distance between corresponding tangent vector and the origin. (3) The bilinear transformation is further determined by the isometric criterion guaranteeing the geodesic distance on high-dimensional SPD manifold as close as possible to the Euclidean distance in the tangent space of low-dimensional SPD manifold. Then, we use it for the DR of original SPD data. Experiments on five commonly used datasets show that RMTSISOM-SPDDR is superior to five advanced SPD data DR algorithms.