Association between Mobility of Residual Left Atrial Thrombus and Stroke Severity in Patients with Nonvalvular Atrial Fibrillation.

BACKGROUND: The differences in the characteristics of ischemic stroke associated with a mobile versus nonmobile residual left atrial thrombus (LAT) are unclear. We investigated whether the mobility of an LAT detected by transthoracic echocardiography is associated with the clinical features of stroke. METHODS: This study included 20 consecutive patients with nonvalvular atrial fibrillation who were admitted to our hospital for treatment of acute ischemic stroke and then found to have an LAT on transthoracic echocardiography. The patients were divided into two groups: those with a mobile LAT (Group M) and those with a nonmobile LAT (Group N). The clinical, neuroradiological, and echocardiographic variables were assessed. RESULTS: The LAT was mobile in 11 patients (Group M) and nonmobile in nine patients (Group N). The median National Institutes of Health Stroke Scale score on admission was higher in Group M than N (17 vs. 7, respectively; p=0.196). Four patients in Group M and one in Group N developed in-hospital stroke recurrence (36% vs. 11%, respectively; p=0.319). The prevalence of large vessel occlusion (15 events in Group M and 10 events in Group N, including in-hospital recurrent events) was significantly higher in Group M than N (73% vs. 30%, respectively; p=0.049), which seemed to lead to poorer functional outcomes in Group M than N (ratio of modified Rankin scale score of 0-2 at discharge: 18% vs. 44%, respectively; p=0.336). CONCLUSIONS: The mobility of LAT may affect stroke severity in patients with nonvalvular atrial fibrillation.

Repeated Ischemic Stroke due to in situ Thrombus at the Middle Cerebral Artery in a Patient with Essential Thrombocythemia.

Essential thrombocythemia (ET) is a myeloproliferative neoplasm that is a rare cause of ischemic stroke. We herein report a 70-year-old man with JAK2 V617F mutation-positive ET who experienced ischemic stroke twice in 1 month due to transient stenosis. In both events, transient stenosis formed at the same curvature of the right middle cerebral artery, and the thrombus disappeared with the initiation of antiplatelet agents. The formation of in situ thrombus at the curvature of the intracranial vessels may be a unique characteristic of JAK2 V617F mutation-positive ET patients.

Particle mobility and macroscopic magnetorheological effects for polyurethane magnetic elastomers.

The relationship between the particle mobility and magnetorheological effect was investigated for polyurethane magnetic elastomers containing carbonyl iron particles with various cross-linking densities or plasticizer concentrations. The storage modulus at 0 mT increased and the on-field modulus at 500 mT decreased with the cross-linking density. The critical magnetic field where the storage modulus starts to rise up increased with the cross-linking density, indicating that the movement of magnetic particles is depressed by the cross-linking points of the polyurethane network. Magnetic elastomers with various plasticizer concentrations revealed that the storage modulus at 0 mT decreased and the on-field modulus at 500 mT increased with the plasticizer concentration. The critical magnetic field decreased with increasing plasticizer concentration, indicating that a dense polyurethane network prevents magnetic particles from moving. It was found that the change in the modulus due to the magnetic field can be scaled by the storage modulus at 0 mT as well as the critical magnetic field. Thus, there is a certain correlation between the macroscopic modulus of elasticity (storage modulus at 0 mT) and the microscopic mobility of magnetic particles reflected in the critical magnetic field.

Effect of Water Absorption on Electric Properties of Temperature-Resistant Polymers.

The effects of water absorption on the electric resistivity and dielectric constant of polyimide (PI) and poly(ethylene terephthalate) (PET) were investigated, and the mechanism of deterioration in electrical insulation properties was discussed. The polyimides are poly(oxydianiline pyromellitimide) (PMDA-ODA) and poly(para-phenylene diamine biphenyltetracarboxydiimide) (BPDA-PDA). These polymer films were immersed in pure water for various immersion times at room temperature, and the water absorption ratio was evaluated. The electric resistance for these films was measured at room temperature using a high-resistance meter, and the dielectric constant at room temperature was measured using an LCR meter in a frequency range of 200 kHz to 2 MHz. The absorption ratios at equilibrium absorption for PMDA-ODA, BPDA-PDA, and PET were 2.7, 2.5, and 0.5%, respectively. The critical volume fraction of the percolation threshold of electric conductivity due to water absorption was 0.034 for both PMDA-ODA and BPDA-PDA. On the other hand, PET did not show a significant decrease in the resistivity. For both PIs and PET, the dielectric constant observed could be explained by a series model of the respective capacitances of pure water and polymer. Actually, the resistivity of samples cut from the edges of the film after water absorption was almost the same value as that in the dry state. These results suggest that the absorbed water molecules are not uniformly dispersed in the film but are localized at the edges of the film even after the absorption equilibrium has been reached.

Deep learning in random neural fields: Numerical experiments via neural tangent kernel.

A biological neural network in the cortex forms a neural field. Neurons in the field have their own receptive fields, and connection weights between two neurons are random but highly correlated when they are in close proximity in receptive fields. In this paper, we investigate such neural fields in a multilayer architecture to investigate the supervised learning of the fields. We empirically compare the performances of our field model with those of randomly connected deep networks. The behavior of a randomly connected network is investigated on the basis of the key idea of the neural tangent kernel regime, a recent development in the machine learning theory of over-parameterized networks; for most randomly connected neural networks, it is shown that global minima always exist in their small neighborhoods. We numerically show that this claim also holds for our neural fields. In more detail, our model has two structures: (i) each neuron in a field has a continuously distributed receptive field, and (ii) the initial connection weights are random but not independent, having correlations when the positions of neurons are close in each layer. We show that such a multilayer neural field is more robust than conventional models when input patterns are deformed by noise disturbances. Moreover, its generalization ability can be slightly superior to that of conventional models.