Experts discuss cutting-edge case studies in AI, dig deeper into the ethical and legal issues, and engage in discussing methods to maximise AI’s potential in the health ecosystem.
Host:
Access to data is directly related to compliance. The first factor you have to consider is data minimisation. It is crucial to maintain a balance in order to effectively use artificial intelligence. An interesting case in France highlights the issue of an organisation intending to use an excessively large database, which led the French advisory authority to intervene. They pointed out that the organisation aimed to conduct a study on prostate cancer but discovered the presence of women's data within the database.
It is far too broad and deemed disproportionate. The advisory authority recommended revising the data set to ensure one has a narrow data set that is aligned with legal requirements. You must take into account the importance of such factors before implementing AI.
Prerequisites encompass the medical device regulation, which encompasses software and apps. The idea of CE marking has been raised, but the situation is more complex. It involves a multi-layered legislative framework.
Consequently, when introducing AI technology and prototypes to the market, it is necessary to provide sufficient time for compliance with all these regulations and rules.
To ensure comprehensive coverage, one must conduct an impact assessment according to Article 55 of the GR. This will help to find the right questions and provide an appropriate checklist. One must address the specific ethics and security problems related to artificial intelligence, such as the inference model and related considerations.
Various organisations, including ICAO in the UK and other ethical and cybersecurity checklists, offer recommendations and factsheets that encompass these things. By referring to these resources, one can obtain a thorough assessment of available AI technology.
The healthcare organisation's AI strategy of transformation should extend beyond the confines of the facility. It is necessary to create a forum where clinicians can use AI to its maximum potential. Formalised checklists and processes can help guide this transformation.
With that in mind, I invite questions from the audience. If anyone has thoughts on the inflection point where the benefits to the organisation outweigh the need for trust.
Audience:
At this stage, we are looking at assessing the use of radiology AI in 10 hospitals. The aim is to scrutinize the organisational benefits of the use of AI and there's quite a deep drift in that.
We are working diligently before presenting the findings. However, we may be slightly behind compared to the UK in terms of assuming additional responsibilities. It may be influenced by the framework of value-based care and bundled payments, which could impact the unconditional adoption of AI.
It is important to invest in the right solutions and seek guidance to avoid the pitfalls that many projects face. For example, international development projects often fail as a result of inadequate funding and planning.
One should consider how to achieve efficiency in terms of configuring already existing products. It's also important to address the challenges of implementing AI in complex environments, including clinical workflows.
Considering the vast amount of data and upcoming regulations, it is important to learn how AI can enhance patient care and patient outcomes.
