Biobanks are biological application systems that are composed of physical samples and databases that centrally store various biological materials and their corresponding information for clinical treatments and life science research. Specific functions of biobank include standardized collection, processing, and storing of physical samples such as tissues, body fluids, and biological derivatives (eg, DNA and RNA), as well as related data storage such as pathological data, clinical diagnosis, treatment records, laboratory omics, ethical approval statements, and informed consent forms [1]. The rudimentary concept of a biobank was born in 1996 [2], and their numbers have increased yearly in the past 20 years with rapid developments in the biomedical industry. Around 636 biobanks opened in the United States in 2013 [3], and 325 institutions in the European-based Biobanking and BioMolecular resources Research Infrastructure were founded until 2018 [4]. However, most biobanks face problems with sustainable development.

As biobanks are an essential link in the precision medicine industry, they require large-scale storage and sample and data management. Significant personnel and funding for equipment, technical services, biomedical consumables, and professional knowledge are necessary for early setup or maintenance. Moreover, biobanks undergo active or passive expansion as the storage sizes increase, which inevitably increases the need for funding [5]. However, as biobanks are not directly involved with production, they often act as an accessory to a specific research group or a collaborative department of public hospitals and have a fuzzy economic operation model [6]. In addition, international rules stipulate that stored human samples should not be used for commercial purposes [7], so establishing a reasonable operational model for biobanks and achieving sustainable development is critical and needs urgent discussion.

Currently, most biobanks in the world, such as the University of California San Francisco AIDS specimen bank, the Wales Cancer Biobank, the Australia Prostate Cancer Bioresource, and the Ontario Cancer Bank [8], recover costs by charging access fees from researchers. Despite this operational model, most biobanks cannot fully recover costs and rely solely on the support of government policies and donations. The Singapore biobank, established in 2002, eventually closed owing to a lack of finances and loss of public and government support [9]. Visibly, programs in which biobanks are developed solely by the government or with funding from a public agency are unrealistic. A path to achieving self-sufficiency is the only solution for long-term survival.

As biobanks form a bridge between clinical medicine and medical research, they must connect with patients who are willing to provide sample data resources and researchers of scientific institutes. Compared to researchers, patients have less initiative and are not as engaged in the entire industrial chain despite them being at the initial link of the chain as the sample and data source. One subjective reason is that patients usually do not fully understand and value the meaning of sample donation. The survey also evidenced that biobank participants are sometimes unaware of the potential risks and benefits of a project [10]; therefore, it may be challenging to convince them to donate.

Biobanks may alleviate funding dilemmas by charging samples and associated data from researchers. Nevertheless, this scheme does not completely resolve the financial burden and can cause other troubles. Relevant data show that sample use in biobanks is <10% [6]. The benefit from only a small number of samples and data may not be enough to afford the long-term management and storage expenses incurred by unused samples and data. In addition, it is difficult for biobanks to define the scope and quality of shared services when providing samples and data to researchers. For example, patient data, particularly those with chronic diseases, often dynamically increases with disease duration. Whereas the availability of biological medical data from biobanks cannot be a “one-time deal,” and determining how the access fee is charged is yet to be discussed. Moreover, security hazards such as data leakage during data sharing or informal access to data may affect the direct benefit of biobanks. It is noteworthy that even when biobanks take considerable economic benefits from operating on patient specimens and data, they can raise ethical and legal issues [11].

Besides ensuring adequate sample quantity, high-quality samples with complete data are prerequisites that guarantee subsequent scientific research output. Many larger biobanks are currently involved in converting physical samples into digital data besides storage management services. Thus, through comprehensive research and analysis of biological samples, genomics, proteomics, transcriptomics, and metabolomics data can be generated, which is convenient for storage and sharing. However, complete sample digitalization is unrealistic, and relying solely on biobanks is impractical. Failure in ensuring data integrity and comprehensiveness also affects subsequent sharing services that make it difficult for researchers to mine for valuable scientific information, in addition to other problems such as data standardization and sharing difficulties.

Throughout the entire industrial chain of translational medicine in biobanks, it is not difficult to find intrinsic motivation weak under its promising prospect. “Biobank” seems to exist merely as “storage” and does not reflect the actual connotation of “bank,” that is, to realize the distribution, value-added, and accelerated turnover of “resource wealth.” Perhaps, biobanks should be positioned not only as pipelines connecting upstream and downstream data but also as self-powered accelerators that vigorously drive research output while robustly accumulating data from clinical samples. Therefore, we introduce the concept of a nonfungible token (NFT) for digital assets that envision a new underlying model to address the existing issues.

In the real world, assets can generally be divided into fungible or nonfungible. Fungible assets such as stocks, game coins, and precious metals tend to have a fixed value that focuses on their quantity rather than their characteristics during any transaction. Unlike fungible assets, nonfungible assets such as houses, buildings, artworks, and game equipment are characterized by uniqueness and irreplaceability (Figure 1).

With the rise and rapid development of blockchain, digital assets have attracted wide attention from academia and industry. These are also classified into fungible and nonfungible assets corresponding to their value in the physical world. The most common fungible digital assets include digital currencies such as Bitcoins and Ethereum. However, nonfungible digital assets include NFTs such as game items and digital artworks [12]. NFTs make it possible to connect the real physical world with the digital network world and have become a core element at the center of an intelligent economy. NFT is a blockchain-based digital asset, which is verifiable, tradable, indivisible, irreplaceable, and unique. Ownership of an NFT can be transferred through intelligent contracts, and this transaction can be recorded by the blockchain; all transaction records can be viewed through any node. The characteristics of blockchain comprise open, transparent, traceable, anticounterfeit, and tamper-proof data that ensure transparency, tamper-proof, and antiduplication of the NFT transaction process [13].

NFT has been successfully applied to many industries. However, it is mainly comprised of digital assets such as artwork, collectibles, video game items, and property rights. Bamakan et al [14] offered a layered conceptual NFT-based framework to create a decentralized market for patent storage and recording that facilitates changing the patent ownership and accelerates their commercialization. In addition, Dos Santos et al [15] proposed the use of NFTs and other tokens to realize transparency of supply chains and trace product supervision from the source and ensure food safety, enhance consumer trust, and realize sustainable agricultural development. With the maturity of blockchain technology and the popularization of its applications, transforming various digital assets and even physical assets into certifications will be a general trend due to the essential advantages of NFT in rights confirmation and transaction.

So far, NFT has not been deployed on a large scale in the field of biomedicine. Ng et al [16] systematically reviewed cutting-edge research on blockchain applications and other digital technologies in medicine and health care. They concluded a strong trend of digitization, such as the application of information technology in the public health response to COVID-19 [17], and blockchain-based patient-centric data management system for managing electronic health records in cancer nursing [18]. Digitalization of medical data will be a good start for NFT construction. NFT relies on a decentralized network that uses advanced encryption to verify the efficacy of transactions. Unlike cryptocurrencies that are seen as currency or digital assets, NFT is equivalent to a digital fingerprint of data or an authenticity certificate. It can hold many files that include transaction IDs and detailed history. Therefore, we can store a large number of biomedical and sample data through an NFT [19]. As envisaged by Skalidis et al [20], tagging and converting patient data (medical information, health status, examination results, and consent forms) into an NFT can enhance privacy and ensure data integrity and confidentiality in clinical practice and research.

NFT technology can also improve the economic model of biobanks. Biobanks should be able to operate much like banks and sustain their operations by earning “interest,” thus reducing their reliance on government funding. To achieve this, we propose the following incentive and economic models.

Biobanks can motivate individuals or research institutions to contribute their biomedical data to the repository through reward mechanisms. When individuals or institutions convert their data into NFTs and store them on the blockchain, they can earn cryptocurrency or other rewards. This incentive model encourages more data contributions, increasing the quantity and diversity of data in the repository.

Researchers or health care institutions seeking access to the repository’s data can pay data access fees. These fees can be determined based on the value and scarcity of the data and can be automatically processed by smart contracts. Some of the data access fees can be allocated to the maintenance and operation of the repository, ensuring its sustainability.

Individuals or institutions storing their data in the repository may need to pay storage fees. This encourages data owners to update and manage their data promptly, avoiding wastage of storage space. Storage fees can also serve as a source of income for the biobank, supporting its operations.

Data quality and credibility are crucial for medical research. The repository can establish reward mechanisms to incentivize data contributors to maintain high-quality, accurate, and complete data. These rewards can be provided in the form of NFTs or cryptocurrencies to encourage data contributors to provide the best data.

Parties participating in the biobank ecosystem, including researchers, health care institutions, developers, and maintainers, can earn ecosystem participation rewards. These rewards can be determined based on their contributions and level of involvement in the ecosystem. This helps build an actively engaged community, driving the growth of the repository.

When data are used multiple times or shared across various research projects, data contributors can earn rewards. This incentivizes the reuse and sharing of data, increasing its value.

These incentive and economic models can be automatically executed on the blockchain through smart contracts, ensuring fairness and transparency. They encourage best practices in data contribution, management, access, and quality, thereby increasing the value and sustainability of biobanks. Moreover, these models contribute to the establishment of a shared and collaborative medical research ecosystem, driving advancements in medical research.

Based on the above assumption, we propose a basic framework of biomedical data based on an NFT, which includes 4 primary levels: application layer, management layer, smart contract layer, and blockchain network layer (Figure 3). The specific structure and implementation of the entire framework are as follows.

By interacting with decenteralized applications, the biomedical data platform allows data providers such as patients and researchers to input or import data via interfaces. After the data are marked as an NFT, its ownership will be transferred to the data demander (schools, hospitals, pharmaceutical enterprises, and other scientific research institutions). Ownership will be eventually tracked through the blockchain.

This will implement user registration and authentication management and use the data processing module to desensitize, clean, and fuse the data to ensure data efficacy and security. Implementation of the interface operation will be done with the NFT smart contract.

Smart contracts generated related to biomedical information as programs running on the blockchain can receive and send transactions. However, they must comply with NFT protocols like Ethereum’s erc721 protocol [25]. After generating a smart contract, the user will obtain a token ID as the NFT ownership certificate. The hash value of the original biomedical information file will be uploaded to the blockchain network for broadcasting, and the metadata of the original file will be stored in the chain [26,27]. The so-called “Mint” process records the ID of an NFT and the address of its owner on the blockchain.

Blockchain systems can be mainly divided into public and private blockchains based on their consensus mechanisms. In public blockchains, any node can participate in the point-to-point network, and it is completely decentralized. In a private blockchain, the read permission can be opened to the public or it can be arbitrarily restricted based on the situation. In addition to these, there is also a semidecentralized alliance chain whose consensus process may be controlled by some designated nodes [28,29]. Public chains with good credit can be selected, such as Ethereum, EOS, Flow, Tezos, and so forth.

Biomedical data are rich in content, and the chain of original files tends to put tremendous pressure on the blockchain network. Thus, the chain of original files is not realistic at present, and off-chain storage can solve the network storage pressure, network pressure, and high maintenance costs [30]. Developers use a function (token uniform resource identifier) to let the app know the location of the metadata of a specific item and establish on-chain and off-chain indexes to realize the association [25,31]. Common off-chain external storage platforms include interPlanetary File System, Filecoin, and Pinata.

Based on the above framework, NFT technology presents potential solutions for the biobank field, enhancing data security, credibility, and management efficiency while also increasing data owners’ participation and control. These technologies hold promise for significant improvements in biobank operations and medical research, providing better data management and use for both patients and researchers.