An improved transformer network for skin cancer classification.

BACKGROUND: Use of artificial intelligence to identify dermoscopic images has brought major breakthroughs in recent years to the early diagnosis and early treatment of skin cancer, the incidence of which is increasing year by year worldwide and poses a great threat to human health. Achievements have been made in the research of skin cancer image classification by using the deep backbone of the convolutional neural network (CNN). This approach, however, only extracts the features of small objects in the image, and cannot locate the important parts. OBJECTIVES: As a result, researchers of the paper turn to vision transformers (VIT) which has demonstrated powerful performance in traditional classification tasks. The self-attention is to improve the value of important features and suppress the features that cause noise. Specifically, an improved transformer network named SkinTrans is proposed. INNOVATIONS: To verify its efficiency, a three step procedure is followed. Firstly, a VIT network is established to verify the effectiveness of SkinTrans in skin cancer classification. Then multi-scale and overlapping sliding windows are used to serialize the image and multi-scale patch embedding is carried out which pay more attention to multi-scale features. Finally, contrastive learning is used which makes the similar data of skin cancer encode similarly so that the encoding results of different data are as different as possible. MAIN RESULTS: The experiment is carried out based on two datasets, namely (1) HAM10000: a large dataset of multi-source dermatoscopic images of common skin cancers; (2)A clinical dataset of skin cancer collected by dermoscopy. The model proposed has achieved 94.3% accuracy on HAM10000 and 94.1% accuracy on our datasets, which verifies the efficiency of SkinTrans. CONCLUSIONS: The transformer network has not only achieved good results in natural language but also achieved ideal results in the field of vision, which also lays a good foundation for skin cancer classification based on multimodal data. This paper is convinced that it will be of interest to dermatologists, clinical researchers, computer scientists and researchers in other related fields, and provide greater convenience for patients.

Value of a Signature of Immune-Related Genes in Predicting the Prognosis of Melanoma and Its Responses to Immune Checkpoint Blocker Therapies.

Melanoma is becoming increasingly common worldwide, with high rates of transformation into malignancy compared to other skin lesions. The prognosis of patients with melanoma at an advanced stage is highly unsatisfying despite the development of immunotherapy, target therapy, or combinative therapy. The major barrier to exploiting immune checkpoint therapies and achieving the best benefits clinically is resistance that can easily develop if regimens are not selected appropriately. In this study, we investigated the possibility of using immune-related genes to predict patient survival and their responses to immune checkpoint blocker therapies with the expression profiles available at The Cancer Genome Atlas (TCGA) Program plus expression data from the Gene Expression Omnibus (GEO) for validation. A five gene signature that is highly correlated with the local infiltration of cytotoxic lymphocytes in the tumor microenvironment was identified, and a scoring model was developed with stepwise regression after multivariate Cox analyses. The score calculated strongly correlates with Breslow depth, and this model effectively predicts the prognosis of patients with melanoma, whether primary or metastasized. It also depicts the heterogenous immune-related nature of melanoma by revealing different predicted responses to immune checkpoint blocker therapies through its correlation to tumor immune dysfunction and exclusion (TIDE) score.

Aberrant expression of the UPF1 RNA surveillance gene disturbs keratinocyte homeostasis by stabilizing AREG.

The up­frameshift suppressor 1 homolog (UPF1) RNA surveillance gene is a core element in the nonsense­mediated RNA decay (NMD) pathway, which impacts a broad spectrum of biological processes in a cell­specific manner. In the present study, the contribution of the NMD pathway to psoriasis lesions and its moderating effects on the biological processes of keratinocytes was reported. Sanger sequencing for skin scales from two patients with psoriasis identified two mRNA mutations (c.2935_2936insA and c.2030­2081del) in the UPF1 gene. The somatic mutants produced truncated UPF1 proteins and perturbed the NMD pathway in cells, leading to the upregulation of NMD substrates. As the most abundant epidermal growth factor receptor ligand in keratinocytes, it was concluded that amphiregulin (AREG) mRNA is a natural NMD substrate, that is dependent on its 3' untranslated region sequence. Perturbed NMD modulated keratinocyte homeostasis in an AREG­dependent but nonidentical manner, which highlighted the unique characteristics of NMD in keratinocytes. By targeting AREG mRNA post­transcriptionally, the UPF1­NMD pathway contributed to an imbalance between proliferation on the one hand, and apoptosis and abnormal differentiation, migration and inflammatory response on the other, in keratinocytes, which indicated a role of the NMD pathway in the full development of keratinocyte­related morbidity and skin diseases.

Repression of M-phase phosphoprotein 8 inhibits melanoma growth and metastasis in vitro and in vivo.

Metastatic melanoma accounts for the majority of skin cancer deaths due to its aggressiveness and high resistance to current therapies. M-phase phosphoprotein 8 (MPP8) has been shown to bind to methylated H3K9 and promote tumor cell motility and invasion. The current study aimed to investigate the role of MPP8 in melanoma growth and metastasis. Our results showed that MMP8 was up-regulated in the metastatic melanoma specimens. Knockdown of MPP8 inhibited melanoma cell growth both in vitro and in vivo. Furthermore, down-regulation of MPP8 induced S-phase cell cycle arrest as well as altered expression of cell cycle-related proteins in melanoma cells. In addition, repression of MPP8 inhibited the migration and invasion of melanoma cells both in vitro and in vivo. Taken together, these data suggest that MPP8 knockdown could inhibit the growth and metastasis of melanoma cells and provide novel therapeutic target for melanoma treatment.