基于皮肤影像大数据的皮肤病人工智能系列产品研发与应用
|
摘要
皮肤病学是基于皮损可视化特征建立的临床二级学科,皮肤影像技术被广泛应用于皮肤病的辅助诊断和治疗评估,可显著提高皮肤科医生的诊断水平和工作效率。人工智能(Artificial Intelligence,AI)是近年来迅速发展的技术和手段,在不同领域均得到应用。"中国人群皮肤影像资源库项目(Chinese Skin Image Database,CSID)"于2017年5月正式成立,已建设成为我国皮肤影像技术研究、教育与应用的共性资源和技术平台。目前,多学科专家基于CSID已完成系列AI产品的研发,并逐步应用到临床工作。皮肤病AI产品在皮肤科中具有广阔的应用前景,将显著提升医生对皮肤疾病的诊疗水平,同时也将改变医生的医疗行为模式,更将赋能基层医院皮肤疾病诊疗。CSID将继续推进皮肤影像技术的普及和皮肤病AI产品在我国的落地化应用,为我国的分级医疗政策和医疗改革做出应有的贡献。
Dermatology is a second-level clinical discipline and based on the lesion visual characteristics, skin-imaging technology is widely used in the auxiliary diagnosis of skin diseases and evaluation of therapeutic effect, which significantly improves the dermatologists' diagnostic level and work efficiency. Artificial intelligence(AI) is an emerging technology and methods in recent years,and has been applied in different fields(including medicine, industry, agriculture, etc.). The Chinese Skin Image Database(CSID)was launched in May 2017, and has developed into a system platform for skin imaging research, education, and application in China.Based on CSID, multidisciplinary experts band together to carry out the research on skin AI, a series of AI products(including portable intelligent dermoscopy, multi-dimensional skin image analysis and management system, Yellow race skin tumor artificial intelligence auxiliary decision-making system-Youzhi AI-1.0, elderly skin tumor remote diagnosis and treatment AI tool, skin-melanonychia AI-1.0, Youzhi AI-2.0, psoriasis management AI) have been completed and gradually applied in clinic, improved the diagnostic level of dermatology in China. Dermatological AI products have broad application prospects in dermatology, which will significantly improve the doctor's diagnosis and treatment level, and will also change the doctor's medical behavior patterns, also enable the primary hospitals to diagnose and treat skin diseases. CSID will continue to promote the spread application of skin imaging technology and the use of Dermatological AI products, and make contribution to the grading medical policy and medical reform in China.
Dermatology is a second-level clinical discipline and based on the lesion visual characteristics, skin-imaging technology is widely used in the auxiliary diagnosis of skin diseases and evaluation of therapeutic effect, which significantly improves the dermatologists' diagnostic level and work efficiency. Artificial intelligence(AI) is an emerging technology and methods in recent years,and has been applied in different fields(including medicine, industry, agriculture, etc.). The Chinese Skin Image Database(CSID)was launched in May 2017, and has developed into a system platform for skin imaging research, education, and application in China.Based on CSID, multidisciplinary experts band together to carry out the research on skin AI, a series of AI products(including portable intelligent dermoscopy, multi-dimensional skin image analysis and management system, Yellow race skin tumor artificial intelligence auxiliary decision-making system-Youzhi AI-1.0, elderly skin tumor remote diagnosis and treatment AI tool, skin-melanonychia AI-1.0, Youzhi AI-2.0, psoriasis management AI) have been completed and gradually applied in clinic, improved the diagnostic level of dermatology in China. Dermatological AI products have broad application prospects in dermatology, which will significantly improve the doctor's diagnosis and treatment level, and will also change the doctor's medical behavior patterns, also enable the primary hospitals to diagnose and treat skin diseases. CSID will continue to promote the spread application of skin imaging technology and the use of Dermatological AI products, and make contribution to the grading medical policy and medical reform in China.
引文
[1]Micali G,Verzi AE,Lacarrubba F.Alternative uses of dermoscopy in daily clinical practice:An update[J].J Am Acad Dermatol,2018,79(6):1117-1132.
[2]Imafuku K,Hata H,Kitamura S,et al.Ultrasound B-mode and elastographic findings of mixed tumour of the skin on the scalp[J].J Eur Acad Dermatol Venereol,2016,30(1):153-5.
[3]Ardigo M,Agozzino M,Franceschini C,et al.Reflectance Confocal Microscopy Algorithms for Inflammatory and Hair Diseases[J].Dermatol Clin,2016,34(4):487-496.
[4]Zheng Y,Lin Y,Cui Y.Teledermatology in China:History,Current Status,and the Next Step[J].J Investig Dermatol Symp Proc,2018,19(2):S71-73.
[5]沈长兵,薛珂,于瑞星,等.推动我国皮肤影像研究、教育与应用的系统平台--中国皮肤影像资源库项目(CSID)[J].皮肤科学通报,2018(2):125-130.
[6]Han SS,Lim W,Kim MS,et al.Interpretation of the Outputs of a Deep Learning Model Trained with a Skin Cancer Dataset[J].J Invest Dermatol,2018,138(10):2275-2277.
[7]Brinker TJ,Hekler A,Utikal JS,et al.Skin Cancer Classification Using Convolutional Neural Networks:Systematic Review[J].J Med Internet Res,2018,20(10):e11936.
[8]Yelamos O,Braun RP,Liopyris K,et al.Usefulness of dermoscopy/dermatoscopy to improve the clinical and histopathologic diagnosis of skin cancers[J].J Am Acad Dermatol,2019,80(2):365-377.
[9]李承旭,沈长兵,崔勇.基于中国人群多维度皮肤影像资源库开展的皮肤病人工智能研究与应用[J].机器人产业,2018(6):94-100.
[10]Alessandrini A,Starace M,Piraccini BM.Dermoscopy in the Evaluation of Nail Disorders[J].Skin Appendage Disord,2017,3(2):70-82.
[11]Shen C,Gao J,Yin X,et al.Association of the late cornified envelope-3 genes with psoriasis and psoriatic arthritis:a systematic review[J].J Genet Genomics,2015,42(2):49-56.
[2]Imafuku K,Hata H,Kitamura S,et al.Ultrasound B-mode and elastographic findings of mixed tumour of the skin on the scalp[J].J Eur Acad Dermatol Venereol,2016,30(1):153-5.
[3]Ardigo M,Agozzino M,Franceschini C,et al.Reflectance Confocal Microscopy Algorithms for Inflammatory and Hair Diseases[J].Dermatol Clin,2016,34(4):487-496.
[4]Zheng Y,Lin Y,Cui Y.Teledermatology in China:History,Current Status,and the Next Step[J].J Investig Dermatol Symp Proc,2018,19(2):S71-73.
[5]沈长兵,薛珂,于瑞星,等.推动我国皮肤影像研究、教育与应用的系统平台--中国皮肤影像资源库项目(CSID)[J].皮肤科学通报,2018(2):125-130.
[6]Han SS,Lim W,Kim MS,et al.Interpretation of the Outputs of a Deep Learning Model Trained with a Skin Cancer Dataset[J].J Invest Dermatol,2018,138(10):2275-2277.
[7]Brinker TJ,Hekler A,Utikal JS,et al.Skin Cancer Classification Using Convolutional Neural Networks:Systematic Review[J].J Med Internet Res,2018,20(10):e11936.
[8]Yelamos O,Braun RP,Liopyris K,et al.Usefulness of dermoscopy/dermatoscopy to improve the clinical and histopathologic diagnosis of skin cancers[J].J Am Acad Dermatol,2019,80(2):365-377.
[9]李承旭,沈长兵,崔勇.基于中国人群多维度皮肤影像资源库开展的皮肤病人工智能研究与应用[J].机器人产业,2018(6):94-100.
[10]Alessandrini A,Starace M,Piraccini BM.Dermoscopy in the Evaluation of Nail Disorders[J].Skin Appendage Disord,2017,3(2):70-82.
[11]Shen C,Gao J,Yin X,et al.Association of the late cornified envelope-3 genes with psoriasis and psoriatic arthritis:a systematic review[J].J Genet Genomics,2015,42(2):49-56.