No abstract

No abstract

Background: It has recently been demonstrated that high-energy diagnostic transthoracic ultrasound and intravenous microbubbles dissolve thrombi (sonothrombolysis) and increase angiographic recanalization rates in patients with ST-segment–elevation myocardial infarction. We aimed to study the effect of sonothrombolysis on the myocardial dynamics and infarct size obtained by real-time myocardial perfusion echocardiography and their value in preventing left ventricular remodeling. Methods: One hundred patients with ST-segment–elevation myocardial infarction were randomized to therapy (50 patients treated with sonothrombolysis and percutaneous coronary intervention) or control (50 patients treated with percutaneous coronary intervention only). Left ventricular volumes, ejection fraction, risk area (before treatment), myocardial perfusion defect over time (infarct size), and global longitudinal strain were determined by quantitative real-time myocardial perfusion echocardiography and speckle tracking echocardiography imaging. Results: Risk area was similar in the control and therapy groups (19.2±10.1% versus 20.7±8.9%; P =0.56) before treatment. The therapy group presented a behavior significantly different than control group over time ( P <0.001). The perfusion defect was smaller in the therapy at 48 to 72 hours even in the subgroup of patients with no recanalization at first angiography (12.9±6.5% therapy versus 18.8±9.9% control; P =0.015). The left ventricular global longitudinal strain was higher in the therapy than control immediately after percutaneous coronary intervention (14.1±4.1% versus 12.0±3.3%; P =0.012), and this difference was maintained until 6 months (17.1±3.5% versus 13.6±3.6%; P <0.001). The only predictor of left ventricular remodeling was treatment with sonothrombolysis: the control group was more likely to exhibit left ventricular remodeling with an odds ratio of 2.79 ([95% CI, 0.13–6.86]; P =0.026). Conclusions: Sonothrombolysis reduces microvascular obstruction and improves myocardial dynamics in patients with ST-segment–elevation myocardial infarction and is an independent predictor of left ventricular remodeling over time.

This paper deals with traffic density reconstruction using measurements from Probe Vehicles (PVs). The main difficulty arises when considering a low penetration rate, meaning that the number of PVs is small compared to the total number of vehicles on the road. Moreover, the formulation assumes noisy measurements and a partially unknown firstorder model. All these considerations make the use of machine learning to reconstruct the state the only applicable solution. We first investigate how the identification and reconstruction processes can be merged and how a sparse dataset can still enable a good identification. Secondly, we propose a pre-training procedure that helps the hyperparameter tuning, preventing the gradient descent algorithm from getting stuck at saddle points. Examples using numerical simulations and the SUMO traffic simulator show that the reconstructions are close to the real density in all cases.