Clinical Data Replication With ML Validation for Offline Trial Integrity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The integrity of data collected during clinical trials for digital therapeutic applications is compromised due to frequent transitions between connected and disconnected states, leading to reduced data validity, inefficient resource allocation, and inaccurate clinical trial results, which affects the effectiveness of digital therapeutics and human-computer interaction.
Innovation Solution
A data replication system that interfaces with multiple data stores and a central repository, using machine learning to validate data integrity by identifying valid data elements and replicating them across these stores, ensuring seamless operation even in offline conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of moving object
If data is accumulated on user device while offline and then sent all at once to server, then data collection continuity is improved, but data integrity is reduced
Solution Approach 1:
The system performs preliminary actions by validating data elements before they are transmitted to the server. The validation server checks data elements while they are still on the user device or in transit, before final storage occurs. This preliminary validation ensures that even if bulk transmission occurs, the data integrity is maintained through pre-checks.
Solution Approach 2:
The system implements feedback mechanisms where validation results are returned to the user device. When data elements are validated, confirmation is provided; when validation fails, the system can request re-collection or correction of data. This feedback loop ensures data integrity is maintained even with offline accumulation and bulk transmission.
2Reliability
If multiple data stores are used for data replication, then data integrity is improved, but device complexity is increased
Solution Approach 1:
The validation server acts as an intermediary between user devices and the central server. It receives data elements from user devices, performs validation, and then facilitates storage in multiple data stores. This intermediary approach enables data replication across multiple stores without requiring each user device to directly manage the complexity of multiple storage locations.
Solution Approach 2:
The system creates copies of validated data elements across multiple data stores. Once data is validated by the validation server, it is replicated to multiple storage locations to ensure redundancy and integrity. This copying mechanism improves reliability while the validation server manages the complexity of coordinating multiple stores.
3Measurement precision
If machine learning validation is implemented, then data validity is improved, but computing resource consumption is increased
Solution Approach 1:
The system applies partial validation by focusing machine learning algorithms on specific data elements that require validation, rather than validating all data uniformly. The validation server selectively applies computational resources to data elements where validation is most critical, reducing overall computing resource consumption while maintaining data validity for key clinical trial metrics.
4Loss of time
If data is transmitted immediately upon connection, then data freshness is improved, but resource allocation efficiency is reduced
Solution Approach 1:
The system performs preliminary validation of data elements before they are fully transmitted and processed by the central server. By validating data elements in advance (while the user device is still connected or during transmission), the system ensures data freshness is maintained while allowing the central server to allocate resources more efficiently, as validation work is distributed to the user device and validation server rather than concentrated on the central server.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided herein are systems and methods for maintaining integrity of data during clinical trials of digital therapeutic applications. A computing system can receive a data element generated based on interactions by a user with a digital therapeutic application during a clinical trial. The computing system can send the data element to each data store of a plurality of data stores. The computing system can access a first data store to retrieve a first instance of the data element. The computing system can identify the first instance of the data element as invalid. The computing system can access a second data store to retrieve a second instance of the data element, responsive to the identifying the first instance of the data element in the first data store as invalid. The computing system can store the second instance of the data element onto a data repository for the clinical trial.