This study accelerates MR cholangiopancreatography (MRCP) acquisitions using deep learning-based (DL) reconstruction at 3T and 0.55T. Thirty healthy volunteers underwent conventional two-fold MRCP scans at field strengths of 3T or 0.55T. We trained a variational network (VN) using retrospectively six-fold undersampled data obtained at 3T. We then evaluated our method against standard techniques such as parallel imaging (PI) and compressed sensing (CS), focusing on peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) as metrics. Furthermore, considering acquiring fully-sampled MRCP is impractical, we added a self-supervised DL reconstruction (SSDU) to the evaluating group. We also tested our method in a prospective accelerated scenario to reflect real-world clinical applications and evaluated its adaptability to MRCP at 0.55T. Our method demonstrated a remarkable reduction of average acquisition time from 599/542 to 255/180 seconds for MRCP at 3T/0.55T. In both retrospective and prospective undersampling scenarios, the PSNR and SSIM of VN were higher than those of PI, CS, and SSDU. At the same time, VN preserved the image quality of undersampled data, i.e., sharpness and the visibility of hepatobiliary ducts. In addition, VN also produced high quality reconstructions at 0.55T resulting in the highest PSNR and SSIM. In summary, VN trained for highly accelerated MRCP allows to reduce the acquisition time by a factor of 2.4/3.0 at 3T/0.55T while maintaining the image quality of the conventional acquisition.

In this paper, we present a deep learning system approach to estimating luminosity, effective temperature, and surface gravity of O-type stars using the optical region of the stellar spectra. In previous work, we compare a set of machine learning and deep learning algorithms in order to establish a reliable way to fit a stellar model using two methods: the classification of the stellar spectra models and the estimation of the physical parameters in a regression-type task. Here we present the process to estimate individual physical parameters from an artificial neural network perspective with the capacity to handle stellar spectra with a low signal-to-noise ratio (S/N), in the $<$20 S/N boundaries. The development of three different recurrent neural network systems, the training process using stellar spectra models, the test over nine different observed stellar spectra, and the comparison with estimations in previous works are presented. Additionally, characterization methods for stellar spectra in order to reduce the dimensionality of the input data for the system and optimize the computational resources are discussed.

As the thickness of the material decreases compared to the other two dimensions, the surface characteristics dominate the bulk properties of the material and then decide its overall physical and chemical behavior [1]. With the advancement of the thin film technology, now it has become possible to create a wide range of variations in the characteristics of the thin-films by controlling the vital parameters of the growth process paving their way of use in the most technologically advanced applications and industries. As in such applications, almost all the properties of a particular thin film depend on its thickness, hence an accurate estimate of the thickness has been one of the most important deciding factor in the application of thin films in industrial sectors. Some examples of such sectors are display industry, semiconductor devices, eye glasses, stents, solar cells, polymer coatings, photoresists, solar panels, LCD, MEMS, thin-film packaging etc. There have been different methods in use for the measurement of the thin film thicknesses. Ion beam analysis, TEM, ellipsometry, surface profilometry etc. are few examples to mention about. In the present work we have proposed to automate the estimation of the thickness of a growing thin-film by applying an artificial neural network (ANN).