Related Experiment Video
Updated: Oct 17, 2026
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
[Design and implementation of a clinical trial image medical record desensitization system based on artificial
Zhili Zhang1, Zhiying Fu2, Min Jiang2
1Key Laboratory of Carcinogenesis and Translational Research(Ministry of Education/Beijing), Information Center, Peking University Cancer Hospital & Institute, Beijing 100142, China.
Objective:
To designe a clinical trial image medical record desensitization system based on artificial intelligence technology and to conduct preliminary applications, achieving the protection of patient privacy and promoting the safe use of clinical image data in clinical trial projects.
Methods:
The clinical trial image record desensitization system based on artificial intelligence technology was first performed to preprocess operations, such as cropping and compression on medical images to initialize the image records. Then, it utilized advanced image recognition, natural language processing, and convolutional neural network technologies to automatically classify (labeled, feature extraction, and text recognition) according to the content of medical documents. It automatically identified sensitive information in the image records of clinical trial patients, including personal identity information such as patient name, identification number, medical record number, phone number, and birthplace. Pixel-level masking or blurring was performed on the identified sensitive information points, and an artificial online desensitization system based on canvas technology was assisted to ensure the security and confidentiality of privacy data.
Results:
The designed clinical trial image record desensitization system exhibited high performance and efficiency in both desensitization effect and processing speed. After desensitizing 100 000 pieces of image record data in this system, a total of 9 000 medical record image samples were randomly selected from the processed medical document types for automated desensitization effect verification. The average accuracy rate of medical document desensitization reached 88.48%, with an accuracy rate of up to 99.40% for serious adverse event medical documents.
Conclusion:
The clinical trial image record desensitization system based on artificial intelligence technology improves accuracy and efficiency in medical data privacy protection. Combined with the artificial online desensitization auxiliary system, it can be applied in production environments, providing reliable technical support for the safe sharing and effective utilization of clinical trial image records.