Yes / Surgical smoke removal algorithms can improve the quality of intra-operative imaging and reduce hazards in image-guided surgery, a highly desirable post-process for many clinical applications. These algorithms also enable effective computer vision tasks for future robotic surgery. In this paper, we present a new unsupervised learning framework for high-quality pixel-wise smoke detection and removal. One of the well recognized grand challenges in using convolutional neural networks (CNNs) for medical image processing is to obtain intra-operative medical imaging datasets for network training and validation, but availability and quality of these datasets are scarce. Our novel training framework does not require ground-truth image pairs. Instead, it learns purely from computer-generated simulation images. This approach opens up new avenues and bridges a substantial gap between conventional non-learning based methods and which requiring prior knowledge gained from extensive training datasets. Inspired by the Generative Adversarial Network (GAN), we have developed a novel generative-collaborative learning scheme that decomposes the de-smoke process into two separate tasks: smoke detection and smoke removal. The detection network is used as prior knowledge, and also as a loss function to maximize its support for training of the smoke removal network. Quantitative and qualitative studies show that the proposed training framework outperforms the state-of-the-art de-smoking approaches including the latest GAN framework (such as PIX2PIX). Although trained on synthetic images, experimental results on clinical images have proved the effectiveness of the proposed network for detecting and removing surgical smoke on both simulated and real-world laparoscopic images. / Research Development Fund Publication Prize Award winner, November 2019.

No description available.

No description available.

Effective light energy conversion into other forms of energy in metal and metal compound nanoparticles has been of great interest in past decades. Being illuminated by incident light, electrons in the nanoparticles can be excited to higher energy states followed by deposition of energy into other molecules around their surface and the lattices in the following relaxation process. Ultrafine nanoparticles are thus preferred in these processes due to their high specific surface areas. Moreover, the portion of excited electrons with higher energies is higher in smaller nanoparticles than in larger ones. However, the overall light power absorbed by nanoparticles is proportional to the square of particle size, which causes the ultrafine nanoparticles not to efficiently absorb the incident light, or to drive further chemical or physical processes.Light antennae materials are usually employed to enhance the light absorption of these ultrafine nanoparticles. Plasmonic nanoparticles, e.g., Ag, Au, Cu, and Al nanoparticles, enhance the light absorption of loaded nanoparticles mainly through strong electromagnetic fields generated near their surfaces and have been proven to be effective light antennae to benefit the light energy conversion of ultrafine nanoparticles. On the other hand, spherical dielectric particles, e.g., silicon dioxide nanospheres, represent a different type of light antennae with the advantages of low cost, simple synthesis, and negligible Ohmic loss when being illuminated. When the sizes of high geometric symmetry dielectric nanospheres are comparable with the wavelength of the incident light, Mie scattering can happen based on the difference in refractive index between the sphere and the surrounding medium, generating size-dependent scattering resonances at various wavelengths. At these wavelengths, strong electric fields can be created on the surface of dielectric spheres to enhance the light absorption of the nanoparticles loaded on the surface. Previous works have shown that silica nanospheres with a diameter of several hundreds of nanometers can effectively enhance the light absorption of ultrafine Pt nanoparticles and benefit photocatalytic reactions, e.g., selective oxidation of benzyl alcohol. Over the past few years, this concept has been broadened to other ultrafine nanoparticles to study their novel photo-to-chemical/physical properties. However, the availability and comprehensive understanding of the optical properties of this class of composite particles still need to be improved. These challenges limit the further development of these composite materials in new light energy conversion processes. This dissertation aims at studying this class of novel ultrafine nanoparticles/dielectric sphere composite particles synthesis and optical properties. In Chapter 2, a synthesis protocol of ultrafine ruthenium oxyhydroxide nanoparticles on the surface of silica nanospheres’ surfaces is introduced. Unlike the traditional synthesis of nanoparticles in solution followed by a loading process, the method developed in this chapter only requires the injection of aqueous ruthenium salt solution into a silica nanosphere dispersion. The obtained ultrafine nanoparticles with sizes of 2-3 nm are characterized to be ruthenium oxyhydroxide (RuOOH) nanoparticles. The silica nanospheres are crucial in stabilizing these ultrafine RuOOH nanoparticles and enhancing their light absorption. Due to the presence of ruthenium-oxygen bonds in the nanoparticles, the absorbed photons are converted to heat and transferred to the surrounding media with a photo-to-thermal conversion efficiency close to the unity. Experimental results have shown that heat can be effectively used in accelerating the reaction rate of selective oxidation of benzyl alcohol by molecular oxygen. Kinetics data also have shown that these ultrafine RuOOH nanoparticles are able to activate molecular oxygen adsorbed on their surfaces, which represents a novel property of these ultrafine RuOOH nanoparticles that is not observed in other traditional ruthenium catalysts. In Chapter 3, a more general synthesis method of ultrafine metal and metal oxyhydroxide nanoparticles on silica nanospheres is developed, inspired by the synthetic route in Chapter 2. Instead of functionalizing silica surfaces with silane agents with amino groups, the silica surfaces are selectively etched by an aqueous base to create a high density of surface hydroxyl groups. These hydroxyl groups can provide basic sites to stabilize metal ions in aqueous dispersion, which are nuclei for the further growth of larger metal oxyhydroxide nanoparticles. In this chapter, more than ten kinds of metal ions are loaded onto silica spheres, forming oxyhydroxide nanoparticles with average sizes below 5 nm. Some oxyhydroxide nanoparticles can be reduced by 5% H2/N2 to form metal nanoparticles with their ultrafine sizes maintained. The synthesis protocol is promising in preparation of bimetallic samples. The composition and optical absorption of all obtained composite particles are analyzed, demonstrating the practicability of utilizing the reported method to prepare high-quality light-absorbing composite particles. In Chapter 4, the optical absorption property of the composite particle is systematically studied. Using ultrafine Pt nanoparticles as the light absorbing material, the light absorptions of composite particles consisting of silica spheres with diameters from 100 to 1100 nm loaded with these Pt nanoparticles are studied. Through the combination of theoretical calculation based on Mie theory and the measured optical absorption spectra, the scattering resonance peaks are successfully located in each sample. It is also found that the photonic crystal effect and the general absorption of Pt nanoparticles can contribute to the light absorption spectra, especially at higher wavelengths. The relationship between the general absorption of Pt nanoparticles and the packing density of the powder is further studied. The successful deconvolution of several components in the absorption spectra can guide the further rational design of composite particles in optical-related applications. In Chapter 5, the composite particle system is further broadened to using high refractive index zinc sulfide nanospheres as a light antenna. The use of a higher refractive index light antenna is promising for obtaining higher light absorption enhancement in loaded ultrafine nanoparticles, even though the sample is dispersed in organic media with a high refractive index as well. After the successful loading of Pt nanoparticles to the surface of silica-coated zinc sulfide nanospheres, a protocol for analyzing their light absorption spectra in organic media is proposed. Size-dependent scattering resonance peaks are observed in bare zinc sulfide nanospheres and can be utilized to enhance the light absorption of Pt nanoparticles, even when the sample is sealed in high refractive index polymeric matrices. The composite particles are further employed in photothermal tests, the results prove that the better light absorption enhancement using zinc sulfide than silica nanospheres. The results introduced in this dissertation represent the first systematic and comprehensive study of ultrafine metal and metal oxyhydroxide nanoparticles loaded on the surface of dielectric light antenna particles. The conclusions open an avenue to further rational design of high-performance light-absorbing composite particles to be used in photo-to-thermal/chemical processes. / Chemistry

Wire electrical discharge machining (WEDM) is a precision machining process that uses electrical discharges struck between an axially moving wire electrode and the workpiece to remove material through melting and vaporization. WEDM is replacing traditional processes like broaching for machining safety-critical components such as the turbine disk in the manufacture of fuel-efficient jet engines. The main issue preventing the more widespread use of WEDM is that due to WEDM being less productive than broaching, it currently requires 6 WEDM machine tools to replace a single broaching machine to maintain the same throughput. The main factor limiting WEDM productivity is wire breakage. To increase the Cutting Rate (CR) more power is required, and increasing power also increases the likelihood of breakage. The goal of this research is to determine whether wires thicker than the conventional 0.25 mm diameter will both optimize the cutting rate and minimize breakage. Thicker wires will allow for an increase in the duty factor, with a significantly decreased incidence of wire breakage. Given that an increased wire diameter also increases the kerf width, this research seeks to identify the optimal wire diameter that maximizes the linear cutting rate. This research concluded that using wire of optimal diameter in WEDM increases the CR by as much as 400%. / Thesis / Master of Science in Mechanical Engineering (MSME) / In order to secure jet engine blades onto the engine, complex features called Firtree Root Forms (FTRF) are used. These features need to be very precisely cut in order for the engine to work at peak efficiency. Currently, industry is using a manufacturing process called broaching to machine these FTRFs, however broaches wear out over time, which causes imprecise cuts. The solution to this is to use Wire Electrical Discharge Machining (WEDM). The problem with WEDM is that it takes on average 6 machines to replace a single broaching machine in terms of productivity. The objective of this project is to increase the cutting speed (and thus productivity) of WEDM, and one of the ways to do that is to increase the electrode wire diameter. This allows for more power to be used in the machining process without the risk of wire breakage, which is a major problem when cutting with WEDM. The research presented in this thesis successfully demonstrates that using thicker wires in WEDM can cut as much as 400% faster than the wires currently in common use in industry.

No description available.

This thesis aims to improve the loading times of web pages by streaming the content in a non-render-blocking way. At the beginning of this thesis, a large-scale analysis was performed, spanning all downloadable pages of the top 10.000 web pages according to the Tranco-list. This analysis aimed to gather data about the render-blocking properties of web page resources, including HTML, JavaScript, and CSS. It further gathered data about code coverage, giving insight into how much of the render-blocking code is actually used. Therefore, the structural optimization potential could be determined. Less render-blocking code will, in turn, lead to faster loading times due to requiring less data to display the page. The analysis showed that there is significant optimization potential left. On average, modern web pages are built with a combined 86.7% of JavaScript and CSS, the rest being HTML. Both JavaScript and CSS are loaded mostly render-blocking, with 91.8% of JavaScript and 89.47% of CSS loaded in this way. Furthermore, only 40.8% of JavaScript and 15.9% of CSS is used until render. This shows that, on average, web pages have significant room for improvement. The concept, which is then developed based on the results of this analysis, aims to load web pages in a new way by streaming all render-blocking content. The related work showed that multiple sub-techniques are required first, which were conceptualized next. First, an optimization and splitting tool for CSS is proposed, called Essential. This is followed by an optimization framework concept for JavaScript, consisting of Waiter and AUTRATAC. Lastly, a backward-compatible approach was developed, which allows for splitting HTML and streaming all content to a client. The evaluation showed that the streamed web page loads significantly faster when comparing FCP, content ”Above-the-Fold,” and total transfer time of all render-blocking resources of the document. For example, the case study test determined that the streamed page could reduce the time until FCP by 83.3% at 2 Mbps and the time until the last render-blocking data is transferred by up to 70.4% at 2 Mbps. Furthermore, existing streaming methods were also compared, determining that WebSockets meets the requirements to stream web page content sufficiently. Lastly, an anonymous online user questionnaire showed that 85% of users preferred this new style of loading pages.

No description available.

No description available.

No description available.