Abstract:

The global incidence of melanoma increased from 2.01/ 100 000 in 1990 to 3.75/100 000 in 2019. The incidence and mortality rate of malignant melanoma in China is lower than that of the global average, but in recent years, its incidence rate has increased (3% to 5%). Malignant melanoma is the fastest growing malignant tumor in China, and new melanoma patients reach 20 000 to 30 000 cases every year. Melanoma is prone to invasion and metastasis, and the 5-year survival rate of stage IV patients is only 4.6%. Early diagnosis and treatment can greatly improve the prognosis of patients with malignant melanoma. At present, the hot spots in the diagnosis of malignant melanoma include visual aid, artificial intelligence-assisted diagnosis, histological diagnosis and differential diagnosis. Visual AIDS can be used to screen for melanoma, and dermoscopy and skin confocal technology (CT) can diagnose melanoma with sensitivity of 61% and 100%, respectively. AI based diagnosis of malignant melanoma is more accurate than that of artificial intelligence alone or doctors, and can improve the diagnosis accuracy of malignant melanoma from 55.6% to 75.0%. With the continuous progress of second-generation sequencing technology in histology diagnosis, there is still a lack of objectively and highly repeatable immunohistochemical markers applicable to all melanomas. However, with the development of melanoma genomics research, BRAF V600E and NRAS mutations have been applied in clinical diagnosis, treatment and prognosis assessment. Malignant melanoma is highly invasive, and many patients are in the advanced stage when diagnosed. Therefore, the diagnosis and treatment of melanoma in China should pay attention to the multidisciplinary diagnosis and treatment team model, and promote the diagnosis and treatment principles and guidelines based on multidisciplinary consensus.

Key words: Melanoma, Dermoscope, Skin CT, Artificial intelligence, Histology, Differential diagnosis