China Conference on Health Information Processing (CHIP2022) released evaluation tasks and continued to accept submissions
The 8th China Conference on Health Information Processing (CHIP2022) is an annual conference on medical, health and biological information processing and data mining and other technologies hosted by the Chinese Chinese Information Society (CIPS) Medical Health and Bioinformation Processing Professional Committee, which is one of the most important academic conferences in the field of processing. The conference focuses on "Real World Research and Digital Therapy", and gathers top medical information processing scholars and medical experts across the countries to discuss the trends and challenges of smart medical development and a new paradigm of medical research and services driven by data and knowledge. The 2022 China Health Information Processing Conference(The official website link is: http://cips-chip.org.cn/2022 )will be held on October 21-23, 2022.
Since 2018, China Health Information Processing Conference has accepted papers every year. The accepted articles will be transferred to SCI indexed journals (JMIR Medical Informatics, etc.) and relevant evaluations of health and medical information technology will be organized. This CHIP2022 technical evaluation announced 5 tasks, including "Text mining task for "Gene-Disease" association semantics, GDAS track", "Medical Causal Entity and Relation Extraction Task", "Extracting Medical Decision Tree from Medical Text", "OCR identification of electronic medical paper documents (ePaper)" and "Clinical Diagnosis Coding Task". At that time, teams that have achieved excellent results in the evaluation will be invited to report and award awards in the evaluation session of the conference. The conference will also provide an official award certificate, and each task has a certain amount of bonus rewards. In addition, the evaluation-winning team will also be invited to write technical papers, which will be published in China core journals under the guidance of the special committee. Researchers in related fields are welcome to participate in the evaluation competition.
Paper submission instructions:
Submission address: https://easychair.org/conferences/?conf=chip2022
Submission deadline: August 10, 2022 August 31, 2022
Acceptance Notification: September 24, 2022
Deadline for final papers: October 8, 2022
Paper Template (Chinese, Using Chinese Journal of Information Templates): http://jcip.cipsc.org.cn/CN/column/column33.shtml
Paper Template (English, Word/Latex): https://www.springer.com/gp/computer-science/lncs/conference-proceedings-guidelines
The specific evaluation tasks are described as follows:
Task 1: Text mining task for "Gene-Disease" association semantics, GDAS track
Task Address: http://cips-chip.org.cn/2022/eval1
Task Introduction: In the massive scientific literature, the association mechanism of "gene-disease" is described by a series of molecular objects such as mutations and genes and their trigger words. Natural language processing provides the possibility to automatically mine this tacit knowledge item, and also provides health and medical information. automated processing solutions. This task includes three sub-tasks: 1. Trigger word entity recognition; 2. Semantic role labeling; 3. Extraction of triples of "gene, regulation type, disease". All data were taken from the AGAC corpus.
Task Organizer:
task organizer: Jingbo Xia,
xiajingbo.math@gmail.com
task contact: Sizhuo Ouyang,
ouyangsizhuo@foxmail.com
Evaluation task address:
http://lit-evi.hzau.edu.cn/AGAC-CHIP2022/
Task 2: Medical Causal Entity and Relation Extraction Task
Task Address: http://cips-chip.org.cn/2022/eval2
Task Introduction: Modern medical care emphasizes interpretability. In diagnosis, treatment and evaluation, doctors are required to focus on the patient and highlight the causal relationship of medical treatment. Therefore, there are a large number of medical questions and answers on the Internet and a large number of causal relationship explanations in knowledge-based texts. While helping patients, it is also of great value to medical search and diagnosis business. From this, it is possible to mine and extract medical causal relationship construction. The causal relationship explanation network builds a medical causal knowledge graph to improve the ability to judge the logic and interpretability of medical results. Following the release of the " Medical Dialogue Clinical Findings Negative and Positive Discrimination Task " at the CHIP2021 conference, the Alibaba Quark medical team released the "Medical Causality Extraction Task" this year.
Task Organizer:
Yixuan Tong, Kangping Yin, Haozi Li, Alibaba quark, China
Mosha Chen/Chuanqi Tan, Alibaba cloud Tianchi, Alibaba DAMO Academy, China
Zhenzhen Lang, Alibaba cloud intelligent Internet (medical expert), China
Buzhou Tang, Harbin Institute of Technology(Shenzhen), PENG CHENG LABORATORY, China
Task 3: Extracting Medical Decision Tree from Medical Text
Task Address: http://cips-chip.org.cn/2022/eval3
Task Introduction: As the core of intelligent medical systems such as auxiliary diagnosis and treatment systems and medical teaching, the acquisition of diagnosis and treatment decision trees often relies on the manual construction of medical experts, which requires a lot of domain knowledge and is time-consuming and labor-intensive. It is very meaningful to automatically extract diagnosis and treatment decision trees (later called Text2DT) from medical textbooks. Clinical diagnosis and treatment can be seen as a process of making judgments based on different conditions and then making different decisions. This clinical diagnosis and treatment process can be modeled as a clinical diagnosis and treatment decision tree. The clinical diagnosis and treatment decision tree is a tree structure composed of conditional nodes and decision nodes. diagnosis and treatment decisions.
Task Organizer:
Xiaoling Wang, East China Normal University, China,
xlwang@cs.ecnu.edu.cn
Wenfeng Li, East China Normal University, China,
51205901094@cs.ecnu.edu.cn
Wei Zhu, East China Normal University, China,
wzhu@stu.ecnu.edu.cn
Yuanbin Wu, East China Normal University, China,
ybwu@cs.ecnu.edu.cn
Wendi Ji, East China Normal University, China,
wdji@cs.ecnu.edu.cn
Buzhou Tang, Harbin Institute of Technology(Shenzhen), PENG CHENG LABORATORY, China
Task 4: OCR identification of electronic medical paper documents (ePaper)
Task Address: http://cips-chip.org.cn/2022/eval4
Task Introduction:
At present, the medical records used in hospitals are still mainly paper, and the information includes: customer information, diagnosis information, medication information, cost information, etc. In the medical industry and insurance industry, this information has high commercial and scientific value, and it is difficult to extract. At present, it still relies on manual input.
With the gradual development and popularization of the application of artificial intelligence technologies such as OCR and NLP in production and life, compared with traditional manual input, the application of OCR and NLP technology can effectively improve work efficiency and reduce the training cost of business personnel. Using OCR and NLP technology to electronically and structure the information on these paper materials has gradually become a hot spot in the current industry.
The data set for this task includes four types of medical record materials: discharge summary, outpatient invoice, drug purchase invoice, and hospitalization invoice. Mainly for needs: pictures of life scenes, extract data, and generate electronic structured data.
Task Organizer:
Lifeng Liu, Xiaobin Zhong, Dejie Chang, Xiaolong Zhao, Tiehu Wang, Jinxin Yang, Beijing global medical assistance, China
Mosha Chen, Special Committee on medical health and biological information processing of Chinese information society, China
Buzhou Tang, Harbin Institute of Technology(Shenzhen), PENG CHENG LABORATORY, China
Task 5: Clinical Diagnosis Coding Task
Task Address: http://cips-chip.org.cn/2022/eval5
Task Introduction:
Disease classification and surgical operation classification and coding are the processing of patient disease diagnosis and treatment information, and are an important part of medical record information management. Medical record coding has become one of the important bases for scientific and information-based management of hospitals. The application of drug monitoring and other aspects are becoming more and more extensive and in-depth.
Among the many classification schemes, the most influential and the most popular in the world is the International Classification of Diseases (ICD). ICD is an international unified disease classification method formulated by WHO, and it is currently a commonly used disease classification method in the world. China has also launched the National Clinical Version 2.0 of the Classification and Code of Diseases and the National Clinical Version 2.0 of the Classification of Surgery Codes, which have been applied in some hospitals.
Task Organizer:
Bo Kang, Yiduyun (Beijing) Technology Co., Ltd, China
bo.kang@yiducloud.cn
Evaluation Chair:
Jianbo Lei, Medical Informatics Center of Peking University, China
Zuofeng Li, Takeda China Innovation Incubator, China
Buzhou Tang, Harbin Institute of Technology(Shenzhen), PENG CHENG LABORATORY, China