Image source: Noway Technology Information Technology Co., Ltd.

    At the first session of the forum, Professor Shen Bairong, Executive Dean of the Disease System Genetics Research Institute of West China Hospital of Sichuan University, Professor Qian Qing, Director of the Medical Informatics Branch of the Chinese Medical Association, Associate Professor Xinghua Mindy Shi of the College of Science and Technology at Temple University, and Professor Liu Qi of the College of Life Science and Technology at Tongji University were invited to attend. Professor Wang Shuang, founder and chairman of Noway Technology, co-hosted the forum with Professor Shen Bairong.

    In his opening remarks, Professor Qian Qing stated that in recent years, data has become an important national strategic resource and factor of production. With the introduction of laws and regulations such as the "Civil Code," "Cybersecurity Law," "Data Security Law," and "Personal Information Protection Law," China has entered an era of strong regulation of data privacy and security. In the process of in-depth integration of information technology and biomedicine, a series of challenges such as data rights confirmation and privacy protection are faced. It is necessary to balance the relationship between privacy protection and data security while promoting the opening and sharing of biomedical data.

    The forum's invitation to heavyweight experts and professors to discuss this topic is very valuable, providing references for common problems encountered in research and work, and fully utilizing the opening and sharing of biomedical data, which is of great significance for serving scientific and technological innovation, improving government public service capabilities, and developing the digital economy.

    Professor Xinghua Mindy Shi delivered a keynote speech on "Trustworthy Biomedical Machine Learning," introducing the application of trustworthy computing and privacy computing in the biomedical field from the perspective of machine learning. Trustworthy computing covers various aspects, including model reliability, robustness, privacy security, and countermeasures against attacks. In addition, data sharing is also a popular research area. In fact, trustworthy computing and data sharing have a two-way relationship; if data security is ensured, it can encourage people to share data. At the same time, if there is a strong willingness or demand for data sharing, it will also accelerate the birth of more effective privacy protection technologies. Therefore, promoting the sharing of biomedical data under the premise of ensuring data and privacy security is a very valuable research issue.

    Next, Professor Xinghua Mindy Shi introduced two large-scale human genome research projects she participated in: The 1000 Genome Project and the Human Genome Structural Variation Consortium. She expressed that some of the issues encountered in this process became the motivation for later research in the fields of machine learning and privacy security. She also introduced the robustness of machine learning in genomics from the perspectives of deep learning and sparse learning. Finally, Professor Xinghua Mindy Shi emphasized the importance of privacy-protected machine learning in biomedicine and expressed her gratitude to Professor Wang Shuang for initiating the iDASH privacy computing competition, providing a platform for peers to exchange privacy computing technologies and applications through healthy competition.

    Professor Liu Qi of the College of Life Science and Technology at Tongji University delivered a keynote speech on "Privacy Computing Case Sharing in the Field of Biomedicine." Professor Liu Qi said that in the process of engaging in precision medicine for complex diseases and drug development, especially in Omics AI research, he gradually realized the importance of privacy protection for biomedical data. In his speech, he introduced how to use specific privacy computing technologies to solve specific problems in the biomedical field and emphasized the integration and sharing of multi-source multi-omics data in biomedical data sharing.

    Professor Liu Qi introduced the integration and sharing of two types of data: biomedical omics data (genome, phenome) and drug data (drug molecules, antibodies), and discussed the application of privacy-protected machine learning in these two aspects and the differences between them. In application, there are differences between drug data and biomedical omics data. In addition to personal privacy and basic data security issues, there are also commercial secrets and patent protection issues. Therefore, when applying various privacy protection technologies, it is also necessary to "prescribe the right remedy for the condition" to better adapt to different scenarios.

03

Looking Forward to the Future

    Professor Wang Shuang said that modern medical research and clinical diagnosis and treatment require a large amount of sharing and joint analysis of biomedical data. Privacy computing can technically ensure the resolution of medical data value circulation and sharing under the premise of data privacy protection and data security, establishing a safe and trustworthy data circulation infrastructure for medical data participants, and promoting the maximization of medical data value utilization. As a startup company with the deepest cultivation in the field of privacy computing + healthcare, we hope to join hands with industry, scientific research groups, colleges and universities, and experts in various fields to build an empowering ecosystem, leading the forefront of privacy computing technology innovation, deeply explore the application potential of privacy computing in the healthcare field, and jointly promote the digital transformation and high-quality development of China's healthcare industry.

    Professor Shen Bairong said that this lecture basically covered the main technologies in the field of data security, such as anonymization, differential privacy, federated learning, etc., and introduced various application scenarios. With the continuous enrichment of data types and the development of technology, the popularity of the internet will promote the superposition of data, bringing challenges to data decryption. I hope everyone will work together to promote the development of this field.

    It is understood that the "Biomedical Data Sharing and Security Series Forum" will continue to focus on topics related to privacy protection and data sharing, such as technologies and models related to privacy and data sharing, legal and compliance issues of medical data, outstanding achievements and cases of biomedical data privacy protection and security sharing, and will hold 1-2 sessions per month in the future.