Verification Begins for Ultra-Long-Term, Large-Capacity Data Storage Service to Support Medical and Research Data Utilization, Including Digital Pathology Images

The Graduate School of Medicine, The University of Tokyo, Department of Human Pathology and Diagnostic Pathology (hereinafter, The University of Tokyo) and NTT East Corporation (hereinafter, NTT East) will begin verification in late July 2026 for a data storage infrastructure that enables safe and efficient long-term storage and utilization of large-capacity medical and research data, such as digital pathology images. This initiative will leverage storage technology provided by International Business Machines Corporation (IBM) and is a new development within the "Remote Bio DX Project*1" that has been underway since 2023.

Through the "Remote Bio DX Project," The University of Tokyo and NTT East have previously verified aspects such as "remote operation of optical microscopes*2" in the fields of clinical pathology, academia, and drug discovery, focusing on operability and usability, and have confirmed the feasibility and utility for their application.

This verification will aim to realize a highly reliable data storage infrastructure that safely and long-term stores large-capacity medical and research data, supporting future AI analysis, re-analysis, and research use. This will be achieved by combining The University of Tokyo's expertise in pathology and actual data such as pathology images, secure access via NTT East's closed network, and IBM's ultra-high-density tape storage, "IBM Storage Deep Archive*3."

*1: NTT East University of Tokyo Lab "Remote Bio DX Project" Introduction Page https://nt-lab.adm.u-tokyo.ac.jp/remote-bioDX/

*2: Industry-Academia Collaboration Succeeds in Realizing Remote Use of Optical Microscopes to Support Research and Clinical Ecosystems https://www.ntt-east.co.jp/release/detail/20260325_01.html

*3: IBM Storage Deep Archive Ultra-high-density tape storage with an S3 Glacier API-compatible interface.

Background and Objectives of This Initiative

With the advancement of generative AI and the evolution of a data-driven society, the need to safely and reliably store data for long-term utilization in the future is increasing across various fields, including medical and research.

In the field of pathology, in particular, many medical institutions face challenges such as increased diagnostic workload and staff shortages, with some operating as "one-person pathology" departments. To address these issues, the adoption and utilization of "digital pathology," which digitizes pathological specimens for remote diagnosis and AI analysis, is progressing. In this context, The University of Tokyo has been a pioneer in the further development of digital pathology in Japan, having introduced a digital pathology education and diagnostic system early on, and has been engaged in research and development for regional medical support through remote pathological diagnosis and the clinical implementation of AI systems to assist pathological diagnosis.

On the other hand, pathology images are extremely high-definition and large-capacity data. As digital pathology becomes more widespread, the amount of data that medical and research institutions need to store is expected to increase further. Therefore, establishing a foundation for safely and efficiently storing large-capacity medical and research data for long-term use, enabling future AI analysis, re-analysis, and research utilization, is becoming crucial.

This initiative will combine The University of Tokyo's expertise in pathology and actual data such as pathology images with NTT East's know-how cultivated in the design, construction, and operation of networks, cloud/data centers, and security. This will enable verification for the safe and efficient storage and utilization of large-capacity medical and research data, including digital pathology images.

Specifically, a secure access infrastructure and data infrastructure utilizing a closed network (not connected to the internet) constructed by NTT East will be combined with IBM's ultra-high-density tape storage, "IBM Storage Deep Archive," to create an environment suitable for long-term storage of large-capacity data.

With technical support from IBM, The University of Tokyo and NTT East aim to realize a highly reliable data storage infrastructure that supports the secure long-term storage and future utilization of medical and research data required in the AI era.

Overview of This Initiative

In this verification, based on the expertise of Professor Tetsuo Ushiku of the Department of Human Pathology and Diagnostic Pathology, Graduate School of Medicine, The University of Tokyo, use cases will be verified using high-definition pathology image data obtained by digitizing pathological specimens. These use cases will assume long-term storage, search and retrieval, and future AI analysis, re-analysis, and research utilization. The goal is to create an environment where large-capacity data, which is increasing with the advancement of digital pathology, can be safely stored over long periods and utilized as needed.

NTT East will leverage its know-how in network, cloud/data center, and security design, construction, and operation to build a secure access infrastructure and data infrastructure using a closed network that does not go through the internet. Furthermore, verification will be conducted on safety, operability, and economic efficiency for use in the medical and research fields, including data encryption, tiered storage based on usage frequency, and compliance with relevant guidelines required for medical information systems.

In addition, NTT East will be the first in Japan to introduce IBM's ultra-high-density tape storage, "IBM Storage Deep Archive," and will verify its reliability, operability, and cost-effectiveness for long-term storage of large-capacity data. This product is a storage solution that can efficiently store and manage large amounts of data. It features cloud-like operations and long-term storage capabilities while reducing the burden of individual operations and specialized knowledge associated with traditional tape usage. It also achieves ultra-high density archiving of up to 61 PB (petabytes)*4 in a standard rack size, along with the low power consumption unique to tape storage and high cyber resilience due to a physical air gap.

*4: 61 PB is equivalent to approximately 61,000 TB (terabytes). If human whole genome analysis data is assumed to be about 100 GB per person, this is enough capacity to store data for about 610,000 people.

Verification Initiative Overview Diagram

Future Development

Based on the knowledge gained from this verification, we will continue to verify the effectiveness of the data storage infrastructure as a foundation for long-term storage of pathology image data, future AI analysis, re-analysis, and research utilization, initially in the digital pathology field. Concurrently, we will consider applying the knowledge gained in the pathology field regarding the storage, search, retrieval, and operation of large-capacity, high-definition image data to the storage and utilization of other diagnostic imaging data and research data in the medical and research fields.

In the future, we aim to realize an ultra-long-term, large-capacity data storage service that supports data utilization in the AI era, with a view to expanding into areas requiring secure long-term storage, such as public documents, video archives, and backup/archiving of critical corporate data.

Through this verification, we will leverage the expertise, technologies, and know-how of each party to contribute to the improvement of medical quality and the advancement of research and development, while supporting the sustainable development of local communities and industries.