Anonymizing Digital Breast Tomosynthesis Data via Secure Hashing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging technologies face challenges in efficiently anonymizing protected health information (PHI) and displaying volumetric images, particularly in scenarios where privacy and legal considerations are paramount, and existing methods for anonymization are cumbersome or lose correlation information between studies.
Innovation Solution
A method that utilizes secure hash algorithms to anonymize PHI by concatenating metadata fields and integrates anonymization into medical imaging systems, allowing for on-the-fly anonymization and secure substitution of patient identifiers, ensuring non-reversibility and institution-aware anonymization without storing mapped IDs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If secure hash algorithms are used to anonymize PHI by concatenating metadata fields, then patient privacy protection is improved, but the complexity of the anonymization process increases
Solution Approach 1:
The patent transforms PHI anonymization by changing the parameter of identifier representation from direct patient identifiers to hash-coded identifiers. The system concatenates metadata fields (patient ID, study ID, exam date) and applies secure hash algorithms to generate unique but non-reversible identifiers, thereby protecting patient privacy while maintaining data usability for research purposes.
Solution Approach 2:
The patent introduces hash-coded identifiers as an intermediary between direct patient identifiers and research data. This intermediary layer allows the system to maintain correlation between studies through consistent hashing of concatenated metadata fields while preventing direct identification of patients, thus resolving the contradiction between privacy protection and data traceability.
2Productivity
If anonymization is integrated into medical imaging systems with on-the-fly processing, then data export efficiency is improved, but the computational load during image processing increases
Solution Approach 1:
The patent applies preliminary action by performing anonymization processing at the time of data acquisition and storage, rather than during data export or analysis. The system generates hash-coded identifiers and applies them to metadata fields immediately when studies are imported into the medical imaging system, so that subsequent data exports and research operations use pre-anonymized data without requiring additional computational resources.
3Loss of information
If hash-coded identifiers are used to maintain correlation between studies, then data traceability is improved, but the risk of re-identification increases
Solution Approach 1:
The patent creates a composite identifier system by concatenating multiple metadata fields (patient ID, study ID, exam date) before applying the hash algorithm. This composite approach ensures that the hash-coded identifier captures sufficient information to maintain correlation between studies while the one-way nature of hashing prevents reverse-engineering to identify individual patients, thus balancing traceability with privacy protection.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
Digital Breast Tomosynthesis allows for the acquisition of volumetric mammography images. The present invention allows for novel ways of viewing such images.In an embodiment a method for displaying volumetric images comprises computing a projection image using a view direction, displaying the projection image and then varying the projection image by varying the view direction. The view direction can be varied based on a periodic continuous mathematical function. A graphics processing unit can be used to compute the projection image and bricking can be used to accelerate the computation of the projection images. In an embodiment of the present invention, users with the appropriate permission can launch a function inside a system in order to anonymize and export generated projection images based on Digital Breast Tomosynthesis, generated projection images based on other volumetric images, other currently loaded studies, or one or more studies identified by a search criteria. The data from the identified studies is then anonymized on the system. In an embodiment of the present invention, the data from selected studies is anonymized on a server, and only then transmitted to another network device or stored to a hard disk or other media.