Search

The increasing market share of light emitting diode (LED) lamps shows that there will be a surplus of LED lamp waste in the future. LED lamps contain base, critical, and precious metals, and these metals are associated with resource depletion and toxicity impacts. To develop effective recycling and disposal programs, updated metal concentration data are crucial. Evaluating the toxic impacts of metals on LED lamp waste is essential for informed decision-making in metal selection and substitution in future LED lamp design. This study aims to evaluate the metal substances in LED lamp models consumed in Thailand, representing the ASEAN region, using standardized leaching tests. It compares the obtained metal concentration data with existing literature and threshold total metal concentration (TTLC) standards to determine the potential hazard status of LED lamp models. Additionally, the study applies the USEtox life cycle impact assessment (LCIA) model to develop a comprehensive toxicological impact data set based on metal concentrations in different LED lamp types. The results showed that copper, aluminum, and iron were the most abundant metals in LED lamps. The average mass concentrations of copper, nickel, silver, lead, and zinc were found to have exceeded the TTLC limits, classifying the waste LED lamps as hazardous. Of the metals found in LED lamp waste, nickel and lead carried the highest cancer risk, zinc and lead the highest noncancer toxicity, and aluminum and iron the highest ecotoxicity. These results indicate that stronger enforcement is needed to reduce toxic impacts. The findings lead to the development of a new data set for recycling/recovery and provide recommendations for making eco-friendly designs in the manufacture of LED lamps. [ABSTRACT FROM AUTHOR]

A segmentation model of the ultrasound (US) images of breast tumors based on virtual agents trained using reinforcement learning (RL) is proposed. The agents, living in the edge map, are able to avoid false boundaries, connect broken parts, and finally, accurately delineate the contour of the tumor. The agents move similarly to robots navigating in the unknown environment with the goal of maximizing the rewards. The individual agent does not know the goal of the entire population. However, since the robots communicate, the model is able to understand the global information and fit the irregular boundaries of complicated objects. Combining the RL with a neural network makes it possible to automatically learn and select the local features. In particular, the agents handle the edge leaks and artifacts typical for the US images. The proposed model outperforms 13 state-of-the-art algorithms, including selected deep learning models and their modifications. [ABSTRACT FROM AUTHOR]