Somatic mutations in facial skin from countries of contrasting skin cancer risk

Semi-supervised skin cancer diagnosis based on self-feedback threshold focal learning

Worldwide, skin cancer prevalence necessitates accurate diagnosis to alleviate public health burdens. Although the application of artificial intelligence in image analysis and pattern recognition has improved the accuracy and efficiency of early skin cancer diagnosis, existing supervised learning methods are limited due to their reliance on a large amount of labeled data. To overcome the limitations of data labeling and enhance the performance of diagnostic models, this study proposes a semi-supervised skin cancer diagnostic model based on Self-feedback Threshold Focal Learning (STFL), capable of utilizing partial labeled and a large scale of unlabeled medical images for training models in unseen scenarios. The proposed model dynamically adjusts the selection threshold of unlabeled samples during training, effectively filtering reliable unlabeled samples and using focal learning to mitigate the impact of class imbalance in further training. The study is experimentally validated on the HAM10000 dataset, which includes images of various types of skin lesions, with experiments conducted across different scales of labeled samples. With just 500 annotated samples, the model demonstrates robust performance (0.77 accuracy, 0.6408 Kappa, 0.77 recall, 0.7426 precision, and 0.7462 F1-score), showcasing its efficiency with limited labeled data. Further, comprehensive testing validates the semi-supervised model's significant advancements in diagnostic accuracy and efficiency, underscoring the value of integrating unlabeled data. This model offers a new perspective on medical image processing and contributes robust scientific support for the early diagnosis and treatment of skin cancer.

Comparison of photodamage in non-pigmented and pigmented human skin equivalents exposed to repeated ultraviolet radiation to investigate the role of melanocytes in skin photoprotection

Introduction

Daily solar ultraviolet (UV) radiation has an important impact on skin health. Understanding the initial events of the UV-induced response is critical to prevent deleterious conditions. However, studies in human volunteers have ethical, technical, and economic implications that make skin equivalents a valuable platform to investigate mechanisms related to UV exposure to the skin. In vitro human skin equivalents can recreate the structure and function of in vivo human skin and represent a valuable tool for academic and industrial applications. Previous studies have utilised non-pigmented full-thickness or pigmented epidermal skin equivalents to investigate skin responses to UV exposure. However, these do not recapitulate the dermal-epidermal crosstalk and the melanocyte role in photoprotection that occurs in vivo. In addition, the UV radiation used in these studies is generally not physiologically representative of real-world UV exposure.

Methods

Well-characterised pigmented and non-pigmented skin equivalents that contain human dermal fibroblasts, endogenous secreted extracellular matrix proteins (ECM) and a well-differentiated and stratified epidermis have been developed. These constructs were exposed to UV radiation for ×5 consecutive days with a physiologically relevant UV dose and subsequently analysed using appropriate end-points to ascertain photodamage to the skin.

Results

We have described that repeated irradiation of full-thickness human skin equivalents in a controlled laboratory environment can recreate UV-associated responses in vitro, mirroring those found in photoexposed native human skin: morphological damage, tanning, alterations in epidermal apoptosis, DNA lesions, proliferation, inflammatory response, and ECM-remodelling.

Discussion

We have found a differential response when using the same UV doses in non-pigmented and pigmented full-thickness skin equivalents, emphasising the role of melanocytes in photoprotection.

The Skin Molecular Ecosystem Holds the Key to Nevogenesis and Melanomagenesis

Early detection of melanoma is critical to good patient outcomes, but we still know little about the mechanisms of early melanoma development. Normal epidermis has many of the sequence variants and genetic architecture disruptions found in both benign nevi, melanomas, and other skin cancers, yet continues to behave more or less normally. One hypothesis is that many melanocytes in this context are "tumor competent" but are regulated by the microenvironment provided by the surrounding keratinocytes to inhibit progress to nevi or melanoma. There is evidence of accumulating disorder in several measures of the genomic and epigenomic landscape from normal skin through nevi to melanoma that may be key to promoting nevogenesis and melanomagenesis.

Characterization of highly active mutational signatures in tumors from a large Chinese population

The majority of mutational signatures have been characterized in tumors from Western countries and the degree to which mutational signatures are similar or different in Eastern populations has not been fully explored. We leveraged a large-scale clinical sequencing cohort of tumors from a Chinese population containing 25 tumor types and found that the highly active mutational signatures were similar to those previously characterized1,2. The aristolochic acid signature SBS22 was observed in four soft tissue sarcomas and the POLE-associated signature SBS10 was observed in a gallbladder carcinoma. In lung adenocarcinoma, the polycyclic aromatic hydrocarbon (PAH) signature SBS4 was significantly higher in males compared to females but not associated with smoking status. The UV-associated signature SBS7 was significantly lower in cutaneous melanomas from the Chinese population compared to a similar American cohort. Overall, these results add to our understanding of the mutational processes that contribute to tumors from the Chinese population.