NLOS Identification in WLANs Using Deep LSTM with CNN Features.

Identifying channel states as line-of-sight or non-line-of-sight helps to optimize location-based services in wireless communications. The received signal strength identification and channel state information are used to estimate channel conditions for orthogonal frequency division multiplexing systems in indoor wireless local area networks. This paper proposes a joint convolutional neural network and recurrent neural network architecture to classify channel conditions. Convolutional neural networks extract the feature from frequency-domain characteristics of channel state information data and recurrent neural networks extract the feature from time-varying characteristics of received signal strength identification and channel state information between packet transmissions. The performance of the proposed methods is verified under indoor propagation environments. Experimental results show that the proposed method has a 2% improvement in classification performance over the conventional recurrent neural network model.

Noninvasive study of coronary microcirculation response to a cold pressor test.

BACKGROUND: The aims of this study were to noninvasively (i) assess the coronary microcirculation changes in response to a cold pressor test (CPT) in control subjects, nondiabetic obese patients and patients with type 2 diabetes and (ii) investigate the response of the coronary microcirculation in patients with diabetes according to the presence or the absence of silent myocardial ischaemia (SMI), asymptomatic coronary stenosis (CS) and left ventricle hypertrophy (LVH). METHODS: The mean left anterior descending coronary flow velocity (mCFV) was measured using transthoracic Doppler before and after a CPT in 16 control subjects, 11 obese and 66 asymptomatic diabetic patients with a high cardiovascular risk. Patients with diabetes were screened for SMI using stress myocardial scintigraphy and/or echocardiography. A coronary angiography was performed in those with SMI. RESULTS: At baseline, pressure-rate product (PRP) was correlated with mCFV (r = 0.23; P < 0.05) and left ventricle mass (r = 0.26; P < 0.05) in the whole population. Changes in PRP and mCFV during CPT were correlated with controls (r = 0.58, P < 0.05), obese (r = 0.75, P < 0.01) and diabetic patients without CS (r = 0.56, P < 0.0001) or without LVH (r = 0.63, P < 0.05) but not in diabetic patients with CS or with LVH. In patients with diabetes, SMI was associated with mCFV changes, independent of other parameters (P < 0.05). CONCLUSION: Transthoracic coronary Doppler allows noninvasive study of changes in the coronary microcirculation during CPT. In asymptomatic patients with type 2 diabetes, this method showed that SMI was associated with mCFV changes during CPT and the presence of CS or LVH was associated with a mismatch between coronary microcirculation and myocardial oxygen demand.

Computational modeling with fluid-structure interaction of the severe m1 stenosis before and after stenting.

PURPOSE: Image-based computational models with fluid-structure interaction (FSI) can be used to perform plaque mechanical analysis in intracranial artery stenosis. We described a process in FSI study applied to symptomatic severe intracranial (M1) stenosis before and after stenting. MATERIALS AND METHODS: Reconstructed 3D angiography in STL format was transferred to Magics for smoothing of vessel surface and trimming of branch vessels and to HyperMesh for generating tetra volume mesh from triangular surface-meshed 3D angiogram. Computational analysis of blood flow in the blood vessels was performed using the commercial finite element software ADINA Ver 8.5. The distribution of wall shear stress (WSS), peak velocity and pressure was analyzed before and after intracranial stenting. RESULTS: The wall shear stress distributions from Computational fluid dynamics (CFD) simulation with rigid wall assumption as well as FSI simulation before and after stenting could be compared. The difference of WSS between rigid wall and compliant wall model both in pre- and post-stent case is only minor except at the stenosis region. These WSS values were greatly reduced after stenting to 15~20 Pa at systole and 3~5 Pa at end-diastole in CFD simulation, which are similar in FSI simulations. CONCLUSION: Our study revealed that FSI simulation before and after intracranial stenting was feasible despite of limited vessel wall dimension and could reveal change of WSS as well as flow velocity and wall pressure.