image: The challenges lie in each perspective of data management from subject recruitment to data analysis in clinical trials. The solutions with different emphases are provided by AI and blockchain technology.
Credit: ©Science China Press
Ensuring Integrity
The framework incorporates a consortium blockchain architecture to establish immutable, cryptographically secured records. It utilizes the SM3 hash algorithm coupled with smart contracts to ensure automated adherence to trial data protection regulations, thereby effectively preventing unauthorized alterations. The system employs on-chain pointers referencing off-chain clinical datasets to balance traceability and storage optimization. Penetration testing confirmed robust resistance to cyberattacks, including SQL injection and cross-site scripting, with blockchain’s tamper-evident design thwarting data falsification.
Enhancing Objectivity
AI-driven tools minimize human bias in data collection and analysis. The DeepControl moodel, which employs the convolutional neural networks (CNNs) based on the EfficientNet architecture, evaluates image quality with 90.8% accuracy, ensuring standardized and objective inputs. Moreover, the DeepGrading model, combining CNNs with long short-term memory (LSTM) networks, automates the disease progression grading over sequential patient visits, achieving mean squared errors of 0.115–0.160. These models adhere strictly to trial protocols, reducing inter-site variability in multicenter settings.
Optimizing Efficiency
The integrated Data Management Web Application (DMWA) streamlines workflows across trial phases. Leveraging blockchain technology, the platform facilitates real-time data synchronization, while AI algorithms substantially diminish the need for manual oversight. Performance tests demonstrated a 35% reduction in response time, capability to accommodate 500 concurrent users, and AI-driven processing tasks completed within 1–3 seconds. In addition, consensus-based smart contracts automate regulatory compliance, and redundant blockchain nodes ensure continuity during network disruptions.
This study introduces a pioneering data management system that synergizes AI with blockchain technology, leveraging its demonstrated robustness in security, objectivity, and efficiency. The framework establishes a paradigm shift in clinical trial data governance, with potential scalability across therapeutic domains and trial phases. Its adoption promises to deliver a resilient, transparent, and auditable infrastructure for biomedical research, empowering clinical researchers to uphold data integrity amid increasingly complex regulatory and methodological demands.
Journal
Science Bulletin