| Project Reference: | ITP/005/23LP |
| Project Title: | AI-aided Toxicological Evaluation for Pharmaceutical Products in Hong Kong SAR |
| Hosting Institution: | The Chinese University of Hong Kong (CUHK) |
| Abstract: | According to the latest requirements by Pharmaceutical Inspection Co-operation Scheme (PIC/S), toxicology evaluation is needed for every pharmaceutical product to give scientific evidence and support for the Cleaning Validation. Currently, there are more than 4000 pharmaceutical products manufactured in Hong Kong and some products may have more than one active pharmaceutical ingredients (API). If we manually evaluate the toxicology of all these APIs, it would be too labor intensive and costly to be finished in a few years. On the other hand, the same APIs shared by many manufacturers, to evaluate their toxicology by each manufacturer individually would lead to a great waste of social resources. Thus, we propose to develop an on-line toxicology evaluation platform with the aid of artificial intelligence (AI) technologies. First, a risk-based decision tree matrix will be established with the help of GMP (Good Manufacturing Practice) experts and local toxicologists and be used to determine those related APIs for each product. Second, we'll use DP2 (Download Pool Version 2), the web scrapping system developed by us, to automatically retrieve the toxicological literature of each API and develop AI models to extract key information required for toxicological evaluation. After integration of all the information, a report of toxicological evaluation of each API will be exported. Finally, a website running on cloud will be established, which make it possible for end-users to use all the AI models to generate toxicology evaluation report on-line. The establishment of this platform will not only greatly reduce the cost of toxicology evaluation, but also is expected to achieve toxicology evaluation of all pharmaceutical products in Hong Kong within two years, to ensure that drugs manufactured in Hong Kong meet the requirements of the new version of PIC/S. |
| Project Coordinator: | Prof Zhong Zuo |
| Approved Funding Amount: | HK$ 2.07M |
| Project Period: | 31 Mar 2023 - 30 Mar 2025 |
