🤖 AI Summary
To address DICOM image privacy leakage risks in the XNAT platform, this work proposes an automated de-identification workflow tailored for research data management. Methodologically, it integrates XNAT’s native API, standardized DICOM parsing, and a configurable rule engine to establish, for the first time, a systematic, multi-scenario-adaptive technical pathway within the XNAT ecosystem. It innovatively incorporates a BERT-NER model fine-tuned for address recognition to enhance anonymization of unstructured textual fields and evaluates performance using the MIDI-B benchmark framework. On the MIDI-B Challenge test set, the workflow achieves an overall de-identification accuracy of 99.61% and a real false-negative rate of only 0.19%, representing a 1.7-percentage-point improvement over the baseline. This work fills a critical gap in systematic, production-ready privacy protection for medical imaging data in XNAT environments and delivers a reusable, empirically validated technical paradigm for compliant research data governance.
📝 Abstract
XNAT is a server-based data management platform widely used in academia for curating large databases of DICOM images for research projects. We describe in detail a deidentification workflow for DICOM data using facilities in XNAT, together with independent tools in the XNAT"ecosystem". We list different contexts in which deidentification might be needed, based on our prior experience. The starting point for participation in the Medical Image De-Identification Benchmark (MIDI-B) challenge was a set of pre-existing local methodologies, which were adapted during the validation phase of the challenge. Our result in the test phase was 97.91%, considerably lower than our peers, due largely to an arcane technical incompatibility of our methodology with the challenge's Synapse platform, which prevented us receiving feedback during the validation phase. Post-submission, additional discrepancy reports from the organisers and via the MIDI-B Continuous Benchmarking facility, enabled us to improve this score significantly to 99.61%. An entirely rule-based approach was shown to be capable of removing all name-related information in the test corpus, but exhibited failures in dealing fully with address data. Initial experiments using published machine-learning models to remove addresses were partially successful but showed the models to be"over-aggressive"on other types of free-text data, leading to a slight overall degradation in performance to 99.54%. Future development will therefore focus on improving address-recognition capabilities, but also on better removal of identifiable data burned into the image pixels. Several technical aspects relating to the"answer key"are still under discussion with the challenge organisers, but we estimate that our percentage of genuine deidentification failures on the MIDI-B test corpus currently stands at 0.19%. (Abridged from original for arXiv submission)