Study On The Standardization Of Neuronal Morphology Dataset

As the basic structural unit of the brain,neuron has become the important research object of brain science.Its diverse morphological structure and complex connection patterns provide a structural basis for the realization of brain functions.With the rapid development of neuronal labeling,optical imaging,digital reconstruction and other research techniques,research methods at the level of single neurons are becoming more and more important in neuroscience.Neuronal morphology data reveals individual spatial structure differences of neurons,it has been an important basis for studying the relationship between neuron classification and brain connection since Cajal.In recent years,it has also become an indispensable material to promote the research of brain-inspired intelligence,and is of great significance for the development of neuroscience and related disciplines.Collecting neuronal morphology data and building a digital database is the basic work of neuroscience.Many research teams have realized the importance of storing and sharing such basic data of neuroscience.In the past ten years,more than 10 research websites and databases have been established,which are playing an important role.As the sources of data have become more diverse and complex,and the volume of data has become larger and larger in recent years,a series of problems related to data standardization have gradually emerged.For example,the data collection methods are not uniform,and the quality is uneven.The data description and coding standards are not unified,and the data sharing is difficult.Therefore,there is an urgent need to study these issues for building a standardized and massive neuronal morphology database.Aiming at the problems of different data acquisition technologies and uneven quality,this paper carried out a standardization study of related issues in the data acquisition process.An applicable standardized processing procedures of neuronal morphology data is proposed.Interactive reconstruction is used to obtain data.Manual intervention strategies are used to improve the accuracy of reconstruction.And the neuronal morphology connection structure and branch redundancy is proposed to correct the result.Based on the high-efficiency,stable and high-quality standardized processing procedures,this paper collected and processed more than 1,000 neuronal morphology data,accumulating research data for subsequent work.In view of the inconsistency of data description,coding,this thesis has designed new metadata standard specification and persistent identifier specification.According to a needs analysis based on a comparison of relevant domestic and foreign standards systems,this paper realizes a unified organization,description and identification for the data.The new metadata specification includes a collection of 28 entities,203 elements,and 31 core metadata items.And for the collection,various semantic definitions and description rules are described.The new persistent identifier specification is mainly consisted of encoding rules,structure,representation,and QR(Quick Response)code association,etc.The uniformity of identification for neuronal morphology data is improved by the new persistent.Based on the above research results,this paper designs a suitable database framework for neuronal morphology research.Furthermore,based on this framework,web interactive sharing of morphology data of more than 1,000 neurons is realized.On this platform,users can effectively manage and share neuronal morphology data.And practical research functions such as query,browsing,and analysis are supplied.The application prospects of related standardization are showed in the construction of neuronal morphology database.In summary,this paper basically achieves the purpose of research.By using the standardization scheme,neuronal morphology data resources can be well unified and integrated.However,as we all known,there are many factors that affect the integration and sharing of scientific research data.It is necessary to advance simultaneously from multiple aspects such as policies,information infrastructure,laws and regulations.So,the island of data information can be broken and the use of scientific research results can be maximized.