The studies showed that green tea and/or some constituents can offer some protection against UV-induced DNA damage in human cell cultures and also in human peripheral blood samples taken post-tea ingestion.

This small case series demonstrates that mTHPC-PDT is a useful initial treatment for VIN III. It is relatively selective, shows good cosmesis and conserves form and function. This is a major advantage over surgery. Repeat treatments are also possible, which is important in a condition such as VIN, which tends to be multifocal. Systemic mTHPC-PDT appears to have an advantage over topical 5-aminolaevulinic acid-PDT as the photosensitizer is distributed widely in areas of disease and consequently identifies foci which may not be apparent clinically but become evident when illuminated.

No abstract

When exposed to UVR, MRC5 fibroblasts incubated with mercuric chloride (0-15 microM) for 1 hour show increased DNA damage (as measured by the comet assay) compared to control cells (UVR irradiated but no mercuric chloride). This demonstrates that mercuric chloride and UVR in combination increase DNA damage in a synergistic manner. This may have implications to those exposed to mercury as it suggests that exposure to mercury in the environment may increase sensitivity to sunlight-induced carcinogenesis.

Early backdoor attacks against machine learning set off an arms race in attack and defence development. Defences have since appeared demonstrating some ability to detect backdoors in models or even remove them. These defences work by inspecting the training data, the model, or the integrity of the training procedure. In this work, we show that backdoors can be added during compilation, circumventing any safeguards in the data preparation and model training stages. As an illustration, the attacker can insert weight-based backdoors during the hardware compilation step that will not be detected by any training or data-preparation process. Next, we demonstrate that some backdoors, such as ImpNet, can only be reliably detected at the stage where they are inserted and removing them anywhere else presents a significant challenge. We conclude that machine-learning model security requires assurance of provenance along the entire technical pipeline, including the data, model architecture, compiler, and hardware specification.