Error Fusion of Hybrid Neural Networks for Mechanical Condition Dynamic Prediction.

It is important for equipment to operate safely and reliably so that the working state of mechanical parts pushes forward an immense influence. Therefore, in order to enhance the dependability and security of mechanical equipment, to accurately predict the changing trend of mechanical components in advance plays a significant role. This paper introduces a novel condition prediction method, named error fusion of hybrid neural networks (EFHNN), by combining the error fusion of multiple sparse auto-encoders with convolutional neural networks for predicting the mechanical condition. First, to improve prediction accuracy, we can use the error fusion of multiple sparse auto-encoders to collect multi-feature information, and obtain a trend curve representing machine condition as well as a threshold line that can indicate the beginning of mechanical failure by computing the square prediction error (SPE). Then, convolutional neural networks predict the state of the machine according to the original data when the SPE value exceeds the threshold line. It can be seen from this result that the EFHNN method in the prediction of mechanical fault time series is available and superior.

Reducing False-Positives in Lung Nodules Detection Using Balanced Datasets.

Malignant pulmonary nodules are one of the main manifestations of lung cancer in early CT image screening. Since lung cancer may have no early obvious symptoms, it is important to develop a computer-aided detection (CAD) system to assist doctors to detect the malignant pulmonary nodules in the early stage of lung cancer CT diagnosis. Due to the recent successful applications of deep learning in image processing, more and more researchers have been trying to apply it to the diagnosis of pulmonary nodules. However, due to the ratio of nodules and non-nodules samples used in the training and testing datasets usually being different from the practical ratio of lung cancer, the CAD classification systems may easily produce higher false-positives while using this imbalanced dataset. This work introduces a filtering step to remove the irrelevant images from the dataset, and the results show that the false-positives can be reduced and the accuracy can be above 98%. There are two steps in nodule detection. Firstly, the images with pulmonary nodules are screened from the whole lung CT images of the patients. Secondly, the exact locations of pulmonary nodules will be detected using Faster R-CNN. Final results show that this method can effectively detect the pulmonary nodules in the CT images and hence potentially assist doctors in the early diagnosis of lung cancer.

Unraveling the Growth Mechanism of Silica Particles in the Stöber Method: In Situ Seeded Growth Model.

In this work, we investigated the kinetic balance between ammonia-catalyzed hydrolysis of tetraethyl orthosilicate (TEOS) and subsequent condensation over the growth of silica particles in the Stöber method. Our results reveal that, at the initial stage, the reaction is dictated by TEOS hydrolysis to form silanol monomers, which is denoted as pathway I and is responsible for nucleation and growth of small silica particles via condensation of neighboring silanol monomers and siloxane network clusters derived thereafter. Afterward, the reaction is dictated by condensation of newly formed silanol monomers onto the earlier formed silica particles, which is denoted as pathway II and is responsible for the enlargement in size of silica particles. When TEOS hydrolysis is significantly promoted, either at high ammonia concentration (≥0.95 M) or at low ammonia concentration in the presence of LiOH as secondary catalyst, temporal separation of pathways I and II makes the Stöber method reminiscent of in situ seeded growth. This knowledge advance enables us not only to reconcile the most prevailing aggregation-only and monomer-addition models in literature into one consistent framework to interpret the Stöber process but also to grow monodisperse silica particles with sizes in the range 15-230 nm simply but precisely regulated by the ammonia concentration with the aid of LiOH.

Facile deposition of continuous gold shells on Tween-20 modified Fe3O4 superparticles.

A facile method to deposit continuous Au shells on Fe3O4 colloidal superparticles was developed by introducing Tween-20 as a surface modification agent to maintain the colloidal stability of the Fe3O4 superparticles and provide nucleation and growth sites for the Au shells. The Fe3O4-Au core-shell particles showed excellent chemical stability, superparamagnetic properties and efficient photothermal conversion performance under laser irradiation at 808 nm, and were expected to be a useful material for biodetection and cancer therapy.

FITC and Ru(phen)32+ co-doped silica particles as visualized ratiometric pH indicator.

The performance of fluorescein isothiocyanate (FITC) and tris(1, 10-phenanathroline) ruthenium ion (Ru(phen)32+) co-doped silica particles as pH indicator was evaluated. The emission intensity ratios of the pH sensitive dye (FITC) and the reference dye (Ru(phen)32+) in the particles were dependent on pH of the environment. The changes in emission intensity ratios of the two dyes under different pH could be measured under single excitation wavelength and readily visualized by naked eye under a 365-nm UV lamp. In particular, such FITC and Ru(phen)32+ co-doped silica particles were identified to show high sensitivity to pH around the pKa of FITC (6.4), making them be potential useful as visualized pH indicator for detection of intracellular pH micro-circumstance.

Multicolor dye-doped silica nanoparticles independent of FRET.

Multicolor particles were prepared by incorporating two dyes, one fluorescent (fluorescein isothiocyanate) and one phosphorescent (tris(1,10-phenanathroline) ruthenium ion), into the silica matrix. Colors of the particles can be easily tuned by either varying the doping ratios of the two dyes or changing the excitation wavelength while fixing the ratios. The multicolor character of the particles is less sensitive to the location of the two dyes in the silica, since the luminescence of the particles is independent of Förster resonance energy transfer (FRET).

Tuning the emission properties of Ru(phen)3(2+) doped silica nanoparticles by changing the addition time of the dye during the Stöber process.

A series of tris(1,10-phenanthroline)ruthenium ion (Ru(phen)(3)(2+)) doped silica nanoparticles were prepared by introducing the dye at different stages of the Stöber process. The emission properties of the doped silica particles were found to be dependent on the time (0-8 h) of the dye introduced into the reaction system. A turnover of the emission properties was identified for the doped silica particles by introducing the dye before and after 3 h of the reaction. Compared to the particles prepared by adding the dye at the beginning of the reaction (0 h doping), the particles prepared by introducing the dye before 3 h of the reaction (3 h doping) showed enhanced emission intensity and blue-shifted emission with the delayed addition time. The particles prepared by introducing the dye during the period of 3-8 h of the reaction showed decreased emission intensity and red-shifted emission with the delayed addition time compared to those prepared by introducing the dye at 3 h of the reaction. The emission intensity of the 3 h doping silica particles was about 3.3 times that of the 0 h doping particles, and the emission maximum shifted from 592 to 575 nm correspondingly. The 8 h doping particles showed emission maximum at 581 nm, and their emission intensity was only 15% of that of the 3 h doping particles. However, both the emission intensity and maximum of the 8 h doping particles would be similar to those of the 3 h doping particles after further deposition of silica protection layer. The switching of the emission properties of the doped silica particles prepared by introducing the dye before and after 3 h is attributed to the suppressed aggregation of the dye molecules and decreased thickness of the silica protection layer, respectively.