3 barriers of Europe

Europe is aware of its potential artificial intelligence (AI) in drug development and life. An example is that last December the European Medicines Agency (EMA, for its acronym in English) published a work plan to guide its use in the regulation of these products. However, the application of the tool brings with it Numerous challenges which the organism has to deal with.

EMA has found that three issues arise as challenges in the application of this tool. This is summarized Understanding AlgorithmsThe dangers of Technical failures and the “broad” effect of the instrument.

“AI and machine learning have the potential to effectively support Data acquisition, transformation, analysis and interpretation Throughout the life cycle of the drug. However, this range of applications brings with it numerous challenges,” explains EMA Medical writing.

Focus on these challenges first Understanding AlgorithmsEspecially its design and possible prejudices. We continue to focus on risk Technical failures

And finally, it will have a “widespread” impact on the adoption of AI in drug development and patient health.

To shed light on this issue, the heads of the EMA and Medicines Agencies (HMAfor its acronym in English) “provide” regulators and other interested actors with reflection, guidance and support in the use of AI.

For example, Europe recently launched a work plan to guide the use of AI in the regulation of medicines, which is valid until 2028. To do so, they relied on four parameters: Guidance, policies and manufacturing base; Tools and Technology AI; support and training and finally, experiment.

Each dimension has established objectives and deadlines. Will continue to support the first of them Development and evolution of AI In the life cycle of drugs; The second seeks to identify and provide a network-wide framework for using tools to increase efficiency, improve understanding and data analysis.

as Collaboration and training, initiatives are taken to continuously develop the capacity of the network. Finally, in experiment

A test cycle will be conducted with a maximum duration of six months.

AI to accelerate drug development

But the documentation about this tool itself has not been published in this regard. From the EMA, a draft reflection document on AI was presented in July to support the development, regulation and safe and effective use of human and veterinary medicines. Once the brief, whose public consultation ends at the end of December, is finalized, guidance on AI “will be provided based on input from interested parties,” the regulatory agency explained.

“The document shows that A A human-centered approach All development and deployment of AI and machine learning should be guided. The use of AI in medicine and in the life cycle must always be in accordance with current legal requirements, guarantee due respect for fundamental rights and take into account responsible ethics,” the EMA reminds.

Therefore, in cases where artificial intelligence or machine learning is used for the development, evolution and monitoring of medicines, the EMA “advises” the developers. Initial regulatory supportThat is, through the qualification of innovative products, development methods or scientific advice.