数字化虚拟人体数据获取的现状和对策

Current problems in the data acquisition of digitized virtual human and the countermeasures

数字化虚拟人的研究方兴未艾,任重道远。完整的数字化虚拟人的宏伟研究规划中,应包括“虚拟可视人”、“虚拟物理人”、“虚拟生理人”和“虚拟智能人”4部分内容。在这个研究领域中,首先要解决的是人体数据的获取。目前,国内外先后已经报道了9个人体数据集。本文就这些数据集的现状进行了评析,重点揭示了这些数据集存在的不足和有待改进之处。本文还结合中国数字化虚拟人研究具体情况,提出了今后人体数据获取的对策性建议:(1)人体数据集的构建必须坚持“质量第一”,要特别注重质量,而不是数量和速度;(2)数据集要符合信息技术需求,要加强对组织器官的特异性标识;(3)数据集要注重发展的前景,应有明确的应用针对性。

As a relatively new field of medical science research that has attracted the attention from worldwide researchers, study of digitized virtual human still awaits long-term dedicated effort for its full development. In the full array of research projects of the integrated Virtual Chinese Human project, virtual visible human, virtual physical human, virtual physiome, and intellectualized virtual human must be included as the four essential constitutional opponents. The primary importance should be given to solving the problems concerning the data acquisition for the dataset of this immense project. Currently 9 virtual human datasets have been established worldwide, which are subjected to critical analyses in the paper with special attention given to the problems in the data storage and the techniques employed, for instance, in these datasets. On the basis of current research status of Virtual Chinese Human project, the authors propose some countermeasures for solving the problems in the data acquisition for the dataset, which include (1) giving the priority to the quality control instead of merely racing for quantity and speed, and (2) improving the setting up of the markers specific for the tissues and organs to meet the requirement from infor- mation technology, (3) with also attention to the development potential of the dataset which should have explicit pertinence to specific actual applications.