登录查看更多内容

What are the most important standards for computer vision systems in healthcare?

由人工智能和领英社区提供技术支持

Computer vision systems are transforming healthcare by enabling faster, more accurate, and more accessible diagnosis and treatment of various conditions. However, developing and deploying such systems also poses significant challenges and risks, such as ensuring data quality, privacy, security, ethics, and compliance. In this article, you will learn about some of the most important standards for computer vision systems in healthcare, and why they are essential for achieving safe and effective outcomes.

此文章中的业界达人

由社区从 3 条内容中精选。了解更多

Samantha Glover

??“?? ?????? ???????? ???? ?????????? ????????????????.” CIO, Mathematician, AI Consultant, Research Scientist, Data…

1 Data standards

Data is the fuel of any computer vision system, and it needs to be of high quality, quantity, and diversity to train and test the models. Data standards are the rules and guidelines that ensure the consistency, interoperability, and usability of data across different sources, formats, and platforms. For example, the DICOM standard defines how medical images and related information are stored, exchanged, and processed. Data standards also help to avoid errors, biases, and redundancies that can compromise the performance and reliability of the computer vision system.

添加您的观点

Samantha Glover

??“?? ?????? ???????? ???? ?????????? ????????????????.” CIO, Mathematician, AI Consultant, Research Scientist, Data Science, ISO - Business Loan Broker
举报内容
Computer vision in healthcare is significant… As a use case, statistics in the US show that there are over 12MM diagnostic errors annually, as well as 800K deaths and disabilities due to misdiagnosis. Every AI system including CV is built on the foundation of data. Critical standards for data in CV are the ability to have interoperability; meaning the ability for different systems to exchange data efficiently. For example… With technical interoperability in healthcare there is a massive network with aligning technologies that require accuracy for collaborative purposes. Different hospitals and healthcare organizations would have to communicate with the highest of standards. We can minimize death, disability and more due to CV. #ai

已翻译

赞

2 Privacy standards

Privacy is a fundamental right of patients and a key responsibility of healthcare providers and developers. Privacy standards are the principles and practices that protect the confidentiality, integrity, and availability of personal and sensitive data, such as medical records, images, and biometrics. For example, the HIPAA standard regulates how health information is collected, used, and disclosed in the US. Privacy standards also help to prevent unauthorized access, misuse, or breach of data that can cause harm, distress, or discrimination to the data subjects.

添加您的观点

Samantha Glover

??“?? ?????? ???????? ???? ?????????? ????????????????.” CIO, Mathematician, AI Consultant, Research Scientist, Data Science, ISO - Business Loan Broker
举报内容
Healthcare is a sensitive topic... In the event of a breach or other scenarios such as discrimination; data can be misused toward unfair practices. There are some studies that show that minority women have a 20-30% chance of being misdiagnosed due a direct and/or indirect association to discrimination in healthcare. We must minimize biases in CV models in order to bridge the gap of healthcare in marginalized communities. In the US we have HIPAA as our primary standard to protect patient data privacy. With any CV application in healthcare it is crucial that the model functions within compliance of HIPAA regulations as unauthorized access and breaches can cause significant harm to the patient. It is a hefty challenge but solvable. #ai

已翻译

赞

3 Security standards

Security is a crucial aspect of any computer vision system, especially in healthcare, where the stakes are high and the threats are diverse. Security standards are the measures and mechanisms that ensure the protection and resilience of the computer vision system and its components, such as data, models, networks, and devices. For example, the ISO/IEC 27000 series defines the best practices for information security management systems. Security standards also help to prevent or mitigate the risks of cyberattacks, sabotage, or theft that can compromise the functionality and availability of the computer vision system.

添加您的观点

4 Ethics standards

Ethics is a complex and evolving domain that reflects the values and norms of society and the stakeholders involved in the computer vision system. Ethics standards are the frameworks and guidelines that ensure the respect, fairness, and accountability of the computer vision system and its impacts, such as the quality of care, the safety of patients, and the trust of users. For example, the IEEE P7000 series defines the ethical principles and processes for the design and development of autonomous and intelligent systems. Ethics standards also help to address or avoid the dilemmas, trade-offs, or controversies that can arise from the use or misuse of the computer vision system.

添加您的观点

5 Compliance standards

Compliance is a legal and regulatory requirement that ensures the adherence of the computer vision system to the applicable laws, rules, and policies of the relevant jurisdictions and authorities. Compliance standards are the criteria and procedures that verify and validate the quality, performance, and safety of the computer vision system and its outcomes. For example, the FDA standard regulates how medical devices, including software as a medical device (SaMD), are classified, approved, and monitored in the US. Compliance standards also help to ensure the transparency, accountability, and responsibility of the computer vision system and its developers and users.

添加您的观点

6 Evaluation standards

Evaluation is a continuous and iterative process that assesses and improves the computer vision system and its results. Evaluation standards are the metrics and methods that measure and compare the accuracy, efficiency, and effectiveness of the computer vision system and its algorithms, models, and applications. For example, the ROC curve and the AUC score are common evaluation standards for binary classification problems. Evaluation standards also help to identify and address the strengths, weaknesses, opportunities, and challenges of the computer vision system and its implementation.

添加您的观点

7 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?

添加您的观点

Samantha Glover

??“?? ?????? ???????? ???? ?????????? ????????????????.” CIO, Mathematician, AI Consultant, Research Scientist, Data Science, ISO - Business Loan Broker
(已编辑)
举报内容
I want to finish my thoughts regarding Computer Vision Systems and the importance of interoperability in Healthcare. As stated previously interoperability gives CV systems the ability to exchange data accurately and efficiently. I mentioned technical interoperability but wanted to discuss semantic interoperability. In this regard many different terms in healthcare may mean different things. It is important that we include data models and vocabularies so that data is understood and interpreted the same way within different systems. A misunderstanding in terms with our CV systems could lead to greater numbers in the misdiagnosis statistics I provided earlier. There are many other forms on interoperability that are also crucial. #ai

已翻译

赞

Artificial Intelligence

+ 关注

给文章评分

我们借助人工智能创建了此文章。您认为这篇文章怎么样？

很棒不太好

举报此文章

查看全部

What are the most important standards for computer vision systems in healthcare?

1

2

3

4

5

6

7

1 Data standards

2 Privacy standards

3 Security standards

4 Ethics standards

5 Compliance standards

6 Evaluation standards

7 Here’s what else to consider

Artificial Intelligence

给文章评分

感谢您的反馈

更多Artificial Intelligence相关文章

更多相关阅读内容

What are the most important standards for computer vision systems in healthcare?

1

2

3

4

5

6

7

1 Data standards

2 Privacy standards

3 Security standards

4 Ethics standards

5 Compliance standards

6 Evaluation standards

7 Here’s what else to consider

Artificial Intelligence

给文章评分

感谢您的反馈

查看其他技能