Chain Bazaar Make it easier for the blockchain to land
Chain Bazaar Make it easier for the blockchain to land
Picture丨From the Internet
Picture丨From the Internet
Picture丨From the Internet
first level title
Summary
Summary
A blockchain can be thought of as a distributed database that allows users to track the provenance of data and retroactively modify records of a given data set. Medical applications of blockchain technology are on the rise. Blockchain technology has many potential applications in medical imaging and is often used to track radiology and clinical data.
Clinical applications of blockchain technology include recording documents provided by different authors, including the contribution of AI algorithms to multi-part reports, recording the use of AI algorithms for diagnosis, the possibility of improving the accessibility of information related to electronic medical records, And users can better control the acquisition of personal health reports.
Applications of blockchain technology in research include better tracking of image data in clinical trials, better delineation of images and annotation data generated in AI algorithm training, thereby improving privacy and fairness, and potentially enabling Imaging data for artificial intelligence has better usability.
This article briefly introduces the basic technology and terminology of blockchain technology, focusing on the potential application of blockchain technology in medical imaging.
key point
key point
Blockchain technology is a distributed ledger technology that allows data to be tracked, recording the origin and changes of all data, including imaging data.
Blockchain technology uses cryptography to achieve data integrity and authenticity, data transparency, immutability, and verifiability. The application of blockchain technology in medicine is being widely promoted, with many potential applications in medical imaging.
preface
Blockchain, as a breakthrough technology, has become familiar to the public due to its wide application in the cryptocurrency market. It has many applications not only in the industrial field, but also in healthcare and medical imaging. In the case of secure storage of medical data using a distributed encrypted database, it is possible to securely store information related to creating, updating, or accessing medical data. information.
Estonia has established a complete healthcare ecosystem based on blockchain technology within a decade. Blockchain allows users (patients, doctors, radiologists and scientists) to control how and by whom this medical data is used. This paper aims to explore the potential application of blockchain technology in medical imaging. The challenges of these applications are illustrated, and some use cases are proposed, including image ownership and tracking, image annotation, and potential applications in artificial intelligence.
To limit the length of this article, we only briefly introduce the basics of blockchain technology.
Blockchain technology can be thought of as a distributed database that keeps track of all changes made to the database. Different from traditional databases, blockchain uses encryption technology to achieve data integrity and authenticity, data transparency, immutability and verifiability, and makes information reliable through a decentralized trust network without using a central, trusted the master copy of . It allows users to trace the provenance of data, as well as trace who has manipulated a given data set in the past, and blockchain may be suitable for the tracking of radioactive data.
Blockchain technology was first proposed in 1991 to verify the authenticity of digital documents through hash functions. The words block and chain were first used in a document by a person or group of people under the pseudonym Satoshi Nakamoto (unidentified). While that document used the terms block and chain at the time, it did not define the term blockchain. The word later appeared in a more informal context. It laid the groundwork for the emergence of the cryptocurrency Bitcoin following the 2007 financial crisis, proposing the basis for a public peer-to-peer electronic money system.
Today, most blockchain projects do not create new blockchain networks, but rely on Ethereum with the ability to execute smart contracts.
Blockchain Technology Basics
secondary title
1. Definition of Blockchain and Distributed Ledger Technology
Blockchain relies on multiple nodes rather than one central control node to store data in a distributed network called a distributed ledger. Data is stored in an unchangeable and untamperable (i.e. write once) manner, which makes illegal modification of data very difficult. Data can only be added to the blockchain, once a block is written it cannot be removed or modified. Thus, the data stored is a growing list of records (blocks), one after the other (chain). Blockchain is a distributed database that permanently stores transaction data and establishes a tamper-proof ledger.
The distributed ledger database is distributed on a peer-to-peer network with multiple node devices, where each node replicates and saves an identical copy of the ledger, and updates itself independently. Its main advantage is that there is no central authority or central server.
Blockchain technology is a form of distributed ledger technology, and not all distributed ledgers use blockchain to provide secure and efficient distributed consensus. However, the structure of a blockchain makes it different from other types of distributed ledgers, in that data in a blockchain is grouped together and organized into blocks, which are linked to each other and secured using cryptography. Blockchain technology is therefore ideal for recording events, managing records, processing transactions, tracking assets and voting
secondary title
2. Comparison between blockchain technology and traditional database technology
Unlike traditional database technology, blockchain technology is decentralized, traceable and immutable.
Decentralization refers to the process of verification, storage, maintenance and transmission of data on the blockchain based on a distributed architecture. In this structure, the trust between distributed nodes is established through mathematical methods, rather than the centralized organization of a single node in traditional database technology.
Immutable means that the tampering of any transaction will generate a different hash value (the hash key links the previous block and points to the next block), which will be detected by all other nodes running the same consensus algorithm. Since the blockchain is a shareable public distributed ledger, stored on thousands of nodes, and continuously synchronized in real time, if the attack is successful, it needs to reach more than 51% of the computing power of the entire network.
secondary title
3. Public chain and private chain
This difference can be compared to the difference between an open-access internet and one that is accessible only with approval. Public blockchains offer lower throughput for reaching consensus due to their complexity and wide distribution. Private blockchains enable greater efficiency. By their nature, cryptocurrencies such as Bitcoin hope to attract the maximum number of participants and become public chain networks, while many enterprise applications use private chains to control access to data stored on the blockchain.
Application of Blockchain in Medical Imaging
secondary title
1. Clinical application
(1) Contributions of different authors to the multi-part report, including the AI system
Reporting of complex radiology studies often includes input from multiple experts, especially in unusual cases, often involving multiple body systems, specialties (eg, cardiac imaging for joint radiologist/cardiologist reports), or modalities ( Such as nuclear medicine physician/radiologist joint report PET/MR).
While a single report is usually synthesized from existing inputs, blockchain technology offers the possibility of determining whose opinion or expertise is responsible for each element of the report, thereby facilitating the determination of when further Possibility of more direct consultation with appropriate contributors as information becomes available. Reports can be constructed in separate blocks without requiring an encrypted blockchain. Blockchain could clarify the exact level of responsibility of each author in the evaluation of medical imaging studies, the opinions of different experts will remain individually signed, and the order of each analysis will be preserved.
(2) Record the application of artificial intelligence algorithms in diagnosis
Like multiple human experts writing a single report, radiology reports may contain a combination of radiologist and AI-generated content, such as structured reports pre-populated by AI algorithms. This may be an increasingly common form of hybrid reporting in the near future. Likewise, blockchain technology can help differentiate between different reporting elements and assign responsibility for incorporating information directly to physicians or physicians relying on AI assistance, while identifying which AI algorithms contributed.
(3) Share clinical data, traceability reports and read reports
Electronic medical records (EMRs) based on blockchain technology, which will automatically store information on the blockchain about who contributed to which part of the EMR, have sparked years of debate. Combined with algorithms designed to display relevant data to remote radiologists, these EMRs will allow radiologists to view data relevant to the case they are working on without having to browse through large parts of the EMR. This in turn may lead to more efficient and better interpretation of radiology studies.
Additionally, if radiologists provide follow-up recommendations or describe how incidental findings will be managed, these items could also be stored and verified on the blockchain. Blockchain also allows storage to identify who has effectively read which parts of these electronic medical records. This could be used in the future to track serendipitous findings. All of these potential uses combined will improve the quality of patient care.
Ensuring user control over data sharing is relevant to many applications, and many data sharing platforms supporting user control have been proposed. More specifically, blockchain technology can be used to put the owners of data stored in electronic medical records over control of their medical data. Patients have the right to share sensitive records with the institutions of their choice to improve healthcare. Build systems that allow clinicians to request patient data and enable patients to grant or revoke access. In a similar fashion, blockchain could also enable patients to control the sharing of their image data.
secondary title
2. Application of blockchain technology in research
(1) Biomarker-derived clinical trial framework
Clinical trials are key to informing changes in clinical practice. Rigorous judging and an absolutely transparent audit trail are necessary conditions for successful and reliable judging. Since imaging is often the cornerstone of identifying disease progression or regression, blockchain technology has the potential to introduce a tamper-resistant mechanism for recording imaging data in clinical trials. This includes all stages of image processing, analysis and quantitative evaluation. Imaging data in trials are often derived from pre-specified imaging protocols, and changes in these protocols or their inaccurate documentation may alter the images and thus their measured output.
The implementation of blockchain eliminates the manipulation of images prior to measurement, ensuring the integrity of images and their measurements. This type of auditing is of particular value when images are uploaded or downloaded between participating sites to perform multiple measurements. Any changes to image settings prior to taking measurements could affect the results, whereas date and time stamping by a blockchain system confirming when and by whom changes were made will avoid inappropriate data manipulation and ensure significant changes traceability. Any data corruption will be identified without the need to manually query the data. In an era where outcomes often depend on the validity of imaging measurements, a reliable method of recording measurement history is critical.
In some specific fields, we can use blockchain technology in an imaging experiment to implement algorithms including lesion segmentation and analysis. Lesion segmentation is traditionally manual, and with the development of technology, it is semi-automatic or even fully automated in more and more cases. Algorithms are often trained using human-segmented datasets annotated by experts to achieve similar results automatically. Documentation of the training set, and its adjustments in the context of experts and software, will provide insight into the basis of machine learning output and explain unexpected variations. Data sets for each block can be assigned by patient visit or by examination set at a specific time point in the trial.
When commercial or academic researchers analyze and report their own findings, there is ample evidence of bias. Blockchain technology applied to imaging in clinical trials will provide provenance-assured datasets to third parties for analysis. This will separate the hypothesis driving the trial from the expected outcome. A particular benefit is the use of datasets in the Imaging Biodatabase for secondary studies, which can ensure the integrity of the primary data, especially when pooling data from multiple trials.
(2) Collect data for scientific research, especially for the training of artificial intelligence algorithms
An interesting application of blockchain is recording who contributed data to the training of AI algorithms, including patients and their data, annotations by radiologists, and industrial partners who developed AI algorithms. And this would allow the distribution of financial rewards to different partners, which in turn could be an incentive for patients and radiologists to participate in the process. Using blockchain technology will also allow patients and radiologists to track data usage, giving them greater control over their files. Experimental benchmark image datasets demonstrate that the accuracy, confidentiality, and fairness requirements of collaborative deep learning can be effectively addressed using a combination of distributed, federated deep learning and blockchain technology.
These properties of blockchain technology also help in providing large amounts of data for AI training. In order to train supervised deep learning networks, one needs to provide as much high-quality data and annotations as possible. If the dataset used for training is not large enough, rare cases will not be reliably detected, creating selection bias that affects the generalizability of AI systems. Since we often lack direct insight into the inner workings of deep learning models, bias can be narrow and even dangerous. The artificial intelligence algorithm based on blockchain technology can not only learn from the shared data of multiple institutions, but also track or evaluate its learning through backtracking or simple playback, thus providing more insight and better understanding of artificial intelligence decision-making. Much human supervision. This is indeed a key value of blockchain technology, similar to the tracking of products from producer to consumer. Annotations of the conditions under which a model was trained undoubtedly provide information about its quality.
Another benefit of blockchain technology is that it can be used for attribution determination of the source of materials for teaching and educational resources. Blockchain technology can ensure the determination of intellectual property rights of educational materials, potentially increasing the willingness of major authors to share materials, so as not to lose credit for intellectual property rights.
(3) Dynamic licensing
Dynamic licensing is a new approach to empowering research partners and promoting active participation in the research process. The use of biological databases in the Dwarna project shows that the use of blockchain technology can give individuals access to information and control over how and where their biological samples and data should be used, and when using blockchain technology, promote The user complies with the erasure mandated by the European Unions General Data Protection Regulation (GDPR).
The current method of storing images in a central database and often using physical media to transmit them not only causes delays in the patient path, but also exposes the data to tampering. Patel has developed a cross-domain image sharing framework that uses blockchain technology as a distributed data store to establish a ledger of radiology research and patient-defined access rights. This framework allows for secure and decentralized sharing of medical data. It also enables patients to effectively own their own image data and control access by healthcare providers.
secondary title
3. Use blockchain technology for management
(1) Learning and tracking learning
Blockchain technology can be used to record and track learning outcomes by using smart contracts on the blockchain to maintain the hash value of learning outcomes and manage access rights. Ocheja et al. describe a learning log blockchain (Boll) that enables learners to transfer learning records from one institution to another in a secure and verifiable format. Nasseem et al. analyzed potential use cases of blockchain deployment in the medical education ecosystem to improve the efficiency, security, functionality, and effectiveness of existing infrastructure. They propose the use of blockchain technology to eliminate the problem of fraudulent academic certification.
(2) Medical device supervision
With the increasing importance and complexity of the medical equipment management system (PMS), the amount of data generated is also increasing. Research shows that a private chain with a proof-of-authority consensus mechanism can provide many advantages to different stakeholders involved in the PMS process, such as providing support for new regulatory initiatives. For the increasingly important and complex medical equipment management system (PMS), more and more data are being generated. Research shows that a private data-authorized blockchain with an authoritative consensus mechanism can offer many advantages to different stakeholders involved in pre-governance processes, such as providing support for new regulatory initiatives.
(3) New business models arising from the use of blockchain technology in radiology
To date, many AI applications have demonstrated their potential for enhancing interpretation of medical imaging. One of the major obstacles to the development of supervised deep learning AI algorithms is the lack of a large number of high-quality annotated images. While the incentive for AI companies is to profit by selling algorithms, the incentive for patients to contribute their medical images, and for radiologists to do the extra work of image annotation, is less straightforward. Using blockchain technology to track who has contributed could help overcome these barriers by providing incentives to those who contribute.
Blockchain technology can also be used for the consensus of annotations. Only when multiple participants reach a consensus on the annotations, the annotations are approved, rewarding those who reach consensus and punishing those who do not reach consensus. Blockchain technology can also help create large open databases with data from many different sources provided by many data contributors. Blockchain technology also helps in access protection of personal health data and protection of data integrity.
first level title
Summarize
Summarize
Blockchain technology is a very powerful technology that facilitates the storage of source data in an immutable manner, providing traceability of all modifications to stored data. Blockchain technology, which provides reliable information on how, who, when and where data was generated, has a wide range of applications in medicine and radiology.
Blockchain technology enhances the potential for patients and radiologists to contribute to medical imaging by allowing them to control data usage rights and data annotation circles. Also ideal for storing patient data for clinical use, the radiology community should participate and collaborate in research on the development and implementation of blockchain technology as it pertains to aspects of patient care.
