On October 13, 2019 MICCAI, the world’s top AI medical imaging academic conference, was opened in Shenzhen. The paper entitled “XiangyaDerm: A Clinical Image Dataset of Asian Race for Skin Disease Aided Diagnosis” by the dermatological AI team led by Prof. Chen Xiang of the Dermatology Department of Xiangya Hospital won the Best Paper Award.
Founded in 2004, MICCAI, the acronym of Medical Image Computing and Computer Assisted Intervention Society, is dedicated to promoting the research, education, and practice of medical image computing and computer-assisted medical intervention including biomedical imaging and robotics. The society regularly organizes high-quality international conferences and releases publications to boost the exchange and the dissemination of advanced and professional knowledge and experience of the world’s major institutes and outstanding scientists, doctors, and educators. MICCAI is the highest academic meeting in the field of computer-assisted diagnosis of medical imaging.
MICCAI LABELS, as a part of MICCAI seminars, aims to study the data scarcity in the development of medical image computing. Data scarcity is definitely the most frustrating obstacle in the development of medical image computing. On the surface, there are two solutions. One is to collect, process, and annotate the large dataset. The other is to develop and use fewer data to achieve the same performance. In its history spanning four years, Labels holds one meeting at the MICCAI conference every year to facilitate the introduction and the discussion in the two fields.
The Asian skin database proposed in the award-winning paper of Xiangya Hospital’s dermatological AI team is constructed by the hospital’s dermatology team with its more than 10 years’ experience in virtue of the hospital’s medical big data platform and (skin disease) Internet hospital platform. The database contains 541 categories of and 107,565 patch images marked by the clinical golden standard. The experiment proved that the most advanced CNN’s classification performance in this dataset, availability as the skin disease diagnosis benchmark dataset, and potency to effectively catalyze the research and the applications of Asian skin disease AI. In addition, the team used a cross-over trial to attest the necessity to establish skin databases for different races in different regions, offering guidance for later product launches and promotions.
It is reported that Xiangya Hospital’s dermatology team was the first to carry out the medico-engineering cross-over study of medicine and the Internet. In 2018, with the approval of the Hunan Provincial Health Commission, the team planned to build the first dermatology Internet hospital in China. Based on this, it has undertaken a series of skin disease medical big data researches at Xiangya Hospital by making use of the scientific research strength of the Key Lab. of Skin Cancer and Psoriasis, Hunan Province, and the Engineering Research Center for Skin Health and Disease of Hunan Province. Since the start of the study, Prof. Chen Xiang led Prof. Zhu Wu, Associate Prof. Su Juan, Associate Prof. Kuang Yehong, and attending physicians Zhao Shuang and Li Fangfang to publish 7 papers in prestigious academic magazines including IEEE Transactions on Industrial Informatics (IF: 7.4,Q1), Journal of the European Academy of Dermatology, and Venereology (IF 5.2,Q1), with 3 papers on the JCR Section I. Many of their academic findings were accepted by 2019 Society for Investigative Dermatology, the world’s top dermatology conference and included by Journal of Investigative Dermatology as conference papers. Meanwhile, the team also cooperated with Prof. Huang Weihong from Mobile Medical Education Department of Xiangya Hospital - China Mobile Joint Laboratory, Prof. Wang Yong Guo Kehua, Xie Bin, and Li Yi from the Schools of Automation/Computer Science and Engineering of Central South University, the leading Internet company Tencent, and Prof. Zuo Ke from National University of Defense Technology. The team also obtained the opportunity to carry out the key project of 2018 key R&D plan of the Ministry of Science and Technology, which was Xiangya Hospital’s first major national project based on intelligent medicine and aimed to develop an AI platform for the whole process of skin disease, and heath clinical diagnosis and treatment.
In response to the call of “Healthy China”, the dermatology team of Xiangya Hospital will continue integrating different sides’ excellent resources and using medico-engineering combination to actively promote the incubation and realization of Xiangya Hospital (Skin Disease) Internet Hospital, build a flagship specialty platform and a national engineering center, and develop skin health and disease AI products with Chinese intellectual property rights that apply to clinical practice to serve people.
