MAPI Launches “MAPI Gateway,” a Medical Imaging Data Anonymization Support System
MAPI Launches “MAPI Gateway,” a Medical Imaging Data Anonymization Support System
~Supporting the reduction of operational burden in medical image anonymization ~
Medical AI Promotion Institute (Headquarters: Chuo-ku, Tokyo; Director: Yuki Shimahara; Representative Director: Sho Kawabe; hereinafter “MAPI”) has launched “MAPI Gateway,” a system that automates the entire workflow from extraction to anonymization and output of medical imaging data through direct integration with PACS. The system is designed to support the reduction of operational burden associated with medical image anonymization.
Current Challenges in Medical Image Anonymization
In recent years, as AI research and clinical studies have advanced, the importance of research utilizing large-scale data has been increasing. However, challenges remain in data utilization, including ensuring data integrity and addressing concerns regarding the handling of personal information.
In addition, preparing training data—such as collection, cleaning, and anonymization—requires significant time and effort, placing a heavy burden on practitioners. Furthermore, with the emergence of facial reconstruction technologies from MRI images, conventional methods such as DICOM tag removal alone are becoming insufficient for personal data protection, increasing the need for anonymization of the image data itself.
Moreover, reliance on manual processes and scripts leads to variability in quality and a tendency for know-how to become person-dependent. Challenges such as insufficient logging of processing details—making audits and ethics review (IRB) responses difficult—as well as the difficulty of selecting appropriate tools, have also become apparent.
Background of MAPI Gateway Development
MAPI has strengths in the anonymization and management of medical data and supports both healthcare institutions that provide data and companies and researchers that utilize such data.
To address these challenges, MAPI developed “MAPI Gateway” (a medical imaging data anonymization support system), which integrates with PACS within healthcare institutions to automate the entire process from data extraction to anonymization and output. The system aims to support the safe utilization of medical data while reducing operational burden at clinical sites.
About MAPI Gateway
MAPI Gateway is an on-premise software solution that connects to PACS within healthcare institutions to automate the acquisition, anonymization, and output of medical imaging data. It operates on a single Mac mini and is designed so that all processing is completed within the hospital network, enabling secure processing without transferring data externally.
Key Features
Automated PACS Integration
Automatically retrieves medical imaging data via Query/Retrieve (Q/R) and supports automated data collection using CSV lists, eliminating the need for manual export operations
DICOM Tag Anonymization
Achieves rule-based, reproducible anonymization using a whitelist approach
Automatic Detection and Removal of Burn-in※1
Detects text embedded within images using AI-OCR and applies masking
De-facing / Re-facing※2
Reduces the risk of facial reconstruction and performs anonymization on the image data itself
Automatic Logging and Traceability
Automatically records all processing steps, enabling traceability.
Guideline Compliance
Enables processing in accordance with the “Guidelines for the Utilization of Medical Digital Data in AI Research and Development”※3
Target Users
ealthcare Information Departments
Departments seeking to standardize and streamline anonymization operations
Physicians and Researchers
Individuals who want to reduce time spent on data preparation and focus on research
Pharmaceutical Companies, AI Companies, and Medical Device Vendors
Organizations requiring reduced risk of residual personal information in received data
Provision Model
On-premise (Mac mini) with a one-time license and two years of maintenance
Cost Structure
No additional subscription fees
Data Security
All data processing is completed within the hospital, with no external data transmission
| Item | Description |
| Supported Modalities (ver.1) | CR, CT, MRI, MG |
| De-facing / Re-facing Support | Head MRI (T1, T2, FLAIR), Head CT |
| Output Format | DICOM / NIfTI |
| Acquisition Criteria | CSV, Date, Modality |
| Interface | Japanese-language Web GUI |
Contact
For inquiries regarding this product or consultation on implementation, please contact:
contact@mapi-jp.org
※1 Text information embedded in images (e.g., patient names, examination details)
※2 De-facing: removal of facial features; Re-facing: reconstruction of facial regions in a non-identifiable manner
※3 Ministry of Health, Labour and Welfare, “Guidelines for the Utilization of Medical Digital Data in AI Research and Development” (March 31, 2024)
※The contents of this release are based on information available at the time of announcement and are subject to change without notice.
Company Overview
Company Name: Medical AI Promotion Institute
Representatives: Yuki Shimabara, Sho Kawabe
Head Office: 4F Life Science Building 9, 12-9 Nihonbashi Odenmacho, Chuo-ku, Tokyo
Established: November 8, 2023
Business: Development and operation of the “Next-Generation Medical Database Business,” which processes and provides medical data
URL: https://mapi-jp.org
Facebook: https://www.facebook.com/iryoaisuishinkiko
X: https://twitter.com/iryoai_mapi
Contact for Media Inquiries
Medical AI Promotion Institute, PR
Email: pr@mapi-jp.org