The rapid advancements in artificial intelligence(AI)have significantly impacted modern healthcare,particularly for skin cancer detection in the field of dermatology.Skin cancer has become a considerable public health...The rapid advancements in artificial intelligence(AI)have significantly impacted modern healthcare,particularly for skin cancer detection in the field of dermatology.Skin cancer has become a considerable public health challenge,highlighting the importance of early detection to improve patient outcomes.AI models have revolutionized skin cancer diagnosis by enhancing the accuracy and efficiency of detecting malignancies in the early stages.This review provides a comprehensive comparison of the latest AI technologies used in skin cancer detection,including segmentation,classification,and reinforcement learning models.It also explores the integration of transformer and multimodal technologies,which have further refined the diagnostic process.Despite the remarkable progress,challenges remain in the practical deployment of these AI models,especially those concerning data diversity,model interpretability,and clinical integration.Future research directions should focus on overcoming these limitations to fully harness AI’s potential in dermatology,aiming to increase diagnostic accuracy,reduce healthcare costs,and ultimately save lives.展开更多
基金supported by the Project of NINGBO Leading Medical&Health Discipline(No.2022-F23),China(No.202002N3190)the Public Welfare Projects of Ningbo,China(No.2022S065).
文摘The rapid advancements in artificial intelligence(AI)have significantly impacted modern healthcare,particularly for skin cancer detection in the field of dermatology.Skin cancer has become a considerable public health challenge,highlighting the importance of early detection to improve patient outcomes.AI models have revolutionized skin cancer diagnosis by enhancing the accuracy and efficiency of detecting malignancies in the early stages.This review provides a comprehensive comparison of the latest AI technologies used in skin cancer detection,including segmentation,classification,and reinforcement learning models.It also explores the integration of transformer and multimodal technologies,which have further refined the diagnostic process.Despite the remarkable progress,challenges remain in the practical deployment of these AI models,especially those concerning data diversity,model interpretability,and clinical integration.Future research directions should focus on overcoming these limitations to fully harness AI’s potential in dermatology,aiming to increase diagnostic accuracy,reduce healthcare costs,and ultimately save lives.
