Subject recruitment is a key component that affects the progress and results of clinical trials, and generally conducted with eligibility criteria (includes inclusion criteria and exclusion criteria). The semantic category analysis of eligibility criteria can help optimizing clinical trials design and building automated patient recruitment system. This study explored the automatic semantic categories classification of Chinese eligibility criteria based on artificial intelligence by academic shared task. We totally collected 38 341 annotated eligibility criteria sentences and predefined 44 semantic categories. A total of 75 teams participated in competition, with 27 teams having submitted system outputs. Based on the results, we found out that most teams adopted mixed models. The mainstream resolution was applying pre-trained language models capable of providing rich semantic representation, which were combined with neural network models and used to fine-tune the models with reference to classifier tasks, and finally improved classification performance could be obtained by ensemble modeling. The best-performing system achieved a macro F1 score of 0.81 by using a pre-trained language model, i.e. bidirectional encoder representations from transformers (BERT) and ensemble modeling. With the error analysis we found out that from the point of data processing steps the data pre-processing and post-processing were very important for classification, while from the point of data volume these categories with less data volume showed lower classification performance. Finally, we hope that this study could provide a valuable dataset and state-of-the-art result for the research of Chinese medical short text classification.
With the development of science and technology, artificial intelligence is gradually integrated into every aspect of daily life and the medical field is no exception. Cardiovascular diseases, as the first killer to global health, is the focus of new technologies and methods. In this study, the application of computer vision, natural language processing, robotics and machine learning in cardiovascular disease studies were reviewed and prospected, in order to promote the development for new technologies and applications in the future.
Objective To develop an artificial intelligence (AI)-driven lung cancer database by structuring and standardizing clinical data, enabling advanced data mining for lung cancer research, and providing high-quality data for real-world studies. Methods Building on the extensive clinical data resources of the Department of Thoracic Surgery at Peking Union Medical College Hospital, this study utilized machine learning techniques, particularly natural language processing (NLP), to automatically process unstructured data from electronic medical records, examination reports, and pathology reports, converting them into structured formats. Data governance and automated cleaning methods were employed to ensure data integrity and consistency. Results As of September 2024, the database included comprehensive data from 18 811 patients, encompassing inpatient and outpatient records, examination and pathology reports, physician orders, and follow-up information, creating a well-structured, multi-dimensional dataset with rich variables. The database’s real-time querying and multi-layer filtering functions enabled researchers to efficiently retrieve study data that meet specific criteria, significantly enhancing data processing speed and advancing research progress. In a real-world application exploring the prognosis of non-small cell lung cancer, the database facilitated the rapid analysis of prognostic factors. Research findings indicated that factors such as tumor staging and comorbidities had a significant impact on patient survival rates, further demonstrating the database’s value in clinical big data mining. Conclusion The AI-driven lung cancer database enhances data management and analysis efficiency, providing strong support for large-scale clinical research, retrospective studies, and disease management. With the ongoing integration of large language models and multi-modal data, the database’s precision and analytical capabilities are expected to improve further, providing stronger support for big data mining and real-world research of lung cancer.