- Artificial Intelligence (AI) -
Revolution of the Statistical Analysis and Reporting Process – Evolution from Automation to RPA and Machine Learning
Dong Fang, Pfizer
It is time to think about what Artificial Intelligence (AI) and Automation can do for your organization and how your organization can use AI to make a positive impact to your customer and employees. In this presentation, the author will give an overview about these new technologies and how they can be related to Statistical Analysis and Reporting. The author will also use the automation of Statistical Analysis Report (SAR) as an example to elaborate its development in three stages: conventional rule-based automation, Robotic Process Automation (RPA), and AI. SAR is a required component of the Clinical Study Report (CSR) package for CFDA submissions. It is mainly developed based on CSR(s), and refers to Statistical Analysis Plan (SAP) and Protocol as appropriate. Creative and broad programming technologies including C#, VBA and XML are used in the implementation of automatic SAR writing. Beyond automating existing processes, RPA can be applied to mimic human manipulating the tool’s user interface instead of relying on routine workers. Another key technology will be performing text analytics from the unstructured text documents, such as the CSR, SAP and Protocol, through natural language processing and machine learning which are key technologies of AI.
Dong Fang is Statistical Programmer and Software Developer at Pfizer. With professional business and technical knowledge as his core skill, Dong has built an internal website that includes several tools to automate and standardize some of the routine work performed by statistical programmers and medical writers. Dong started the development of innovative tools in 2012 when he was still a graduate student in a joint graduate program between Pfizer and Fudan University. After joining Pfizer as a Statistical Programmer, Dong has been actively involved in Pfizer innovation initiatives covering Robotic Process Automation (RPA) and Artificial Intelligence (AI).
- Robotics -
CyberKnife – Precision Robotic Treatment as Individual as Every Patient
Zeo Zhang（张旻佳), Clinical Marketing Manager, Accuray APAC
What is CyberKnife , and How it works, and what is the unique benefit for the Radiotherapy treatment. The CyberKnife System, the premier solution for full-body robotic radiosurgery, extending its accuracy and precision to radiation therapy – allowing the very best treatment for each of the patients, with confidence and without compromise. Enable high definition radiotherapy anywhere in the body with the widest range of motion of the industry. Deliver radiosurgery (SRS) and stereotactic body radiotherapy (SBRT) with end-to-end sub-millimeter accuracy. Continually monitor patient and tumor motion to confidently deliver the prescribed dose while sparing surrounding critical structures and healthy tissue.
- Big Data -
Causal analysis of electronic health record data and methodological challenges 电子健康档案数据的统计因果推断及其方法挑战
Bo Fu 傅博, School of Data Science, Fudan University
Dr. Fu will introduce a motivating example of causal mediation analysis of socioeconomic status, birth weight and infant mortality using national administrative data from Scotland and Denmark and discuss methodological challenges in analyzing administrative data and electronic health record data. Existing methods for addressing missing data, complex confounding and unmeasured confounding for causal inference will be reviewed.
傅博，教授，博士生导师，上海市“千人计划”特聘专家。2017年回国前先后在南洋理工大学数学系、曼彻斯特大学医学院、伦敦大学学院医学院任正式教职，在剑桥大学、英国行政数据研究中心和英国健康大数据研究所兼职。主要研究领域是数据科学与统计方法、公共健康、医疗大数据、社会研究、行政数据等
Make Data Preparation More Enjoyable and Efficient
Wenjun Bao, JMP, SAS Institute Inc.
Big data, machine learning, deep learning, and artificial intelligence have become the prestigious technologies that attract almost every industry and field’s attention. To take advantage of these techniques, one unavoidable, tedious and time-consuming step is data preparation, which includes data cleaning and organizing. According to a Forbes survey in 2016, data cleaning and organizing were the least enjoyable and most time consumed tasks for data scientists. This presentation will reveal efficient ways to conduct data preparation through a visualizing statistical tool called JMP. We will demonstrate common and complex data preparation tasks such as data format conversion; outlier investigation; missing values and pattern processing; data modification such as recoding, transformation, concatenation and transpose; and multiple table/Excel sheet join and unification. The presentation will show that data cleaning and organizing can be enjoyable and very efficient, using various embedded formula, functions and visualization capabilities.
Dr. Wenjun Bao is a Chief Scientist for JMP Life Sciences that developed JMP Clinical and JMP Genomics. She received her B.S from China Pharmaceutical University and Ph.D. in Biochemistry from Oregon Graduate Institute of Oregon Health & Science University. Before joining SAS, Dr. Bao was an IRTA fellow at the National Institute of Health (NIH), a faculty at Duke University, and a scientist at the US Environmental Protection Agency. She has rich and diverse experiences in clinical, biochemistry, molecular biology and bioinformatics researches. Dr. Bao has been a research grant review committee member for NIH since 2005 and research advisors for scientists in a number of research projects at Universities and government agencies. She also has multiple publications in the peer-reviewed journals.
