Could Linguistic AI restrict EU MDR translation challenges?

Eric Jacobs 18 Feb 2020 5 min read

With the vast amount of new content requiring translation under the EU Medical Device Regulation (MDR), medical device companies are scrambling to establish a strategy that would lead them to compliance by the May 2020 deadline. A recent KPMG survey reported that only 28% of respondents plan to declare MDR compliance by May 2020 and 66% have not started planning for the long-term organizational impacts of EU MDR. One of the most pressing challenges for medical device organizations is managing language requirements under the new regulation.

In this second blog of our three part series, EU MDR: 3 Best Practices for Multilingual Compliance, we will discuss how integrating Linguistic Artificial Intelligence (AI) into your overall MDR translation strategy could not only speed up the process, but also reduce costs without impacting quality.

What does Linguistic AI mean for most medical device organizations?

AI is undeniably changing the healthcare world, today diabetic patients are using voice assistants to manage their treatment, while virtual doctor apps are relying on Deep Learning to deliver personalized care. Unsurprisingly AI has made significant advancements into the medical devices sphere, from diagnostic technologies to therapeutic applications and robotics.

However, despite widely-discussed advances in Linguistic AI, adoption by the Medical Device industry remains marginal, even though other highly regulated industries are seeing technologies like AI and machine translation (MT) as viable and effective solutions to address global content challenges. Additionally, with these technologies rapidly progressing and many new use cases flourishing, the medical device industry could be missing an opportunity to positively impact their EU MDR translation timelines and budgets.

Sacrificing quality for speed?

One of the biggest perceived drawbacks to implementing a machine-first approach is the quality of the translation – a machine can’t “talk” like a human. Machines are trained. When we hear the term “machine learning” – we think of unsupervised processes where machines get better without any human intervention. That’s only partially true. Machines learn from data and humans have to either kick-start the process or monitor the process such that bias isn’t introduced into the system. By combining Linguistic AI with human post-editing capabilities, the expertise and skillset of the translators ensure that sensitive, life-saving medical content is accurate, respects guidelines and other linguistic preferences.

How secure is Linguistic AI?

Medical device manufacturers may also have security risk and data breach concerns due to the sensitive data these organizations handle. Unfortunately they should be concerned, as online machine translation tools do not provide adequate data security and tailoring controls for the specific needs of medical device organizations. Furthermore once something is translated using these online translation tools that content is potentially available in the public domain. If data privacy and security is a top priority, you should only be implementing enterprise grade solutions with firewall protections.

Not all Linguistic AI use cases are equal

The quality, integrity and volume of data will pretty much dictate if Linguistic AI is a viable approach for your specific use case, completing a detailed analysis should be one of your first steps. Additionally certain areas of content given their structure tend to lend themselves better to this approach, however looking beyond the usual suspects (online help, etc.) would be a wise decision, as we have seen customer success across areas as diverse as labeling and safety.

SDL Linguistic AI ™ (Hai) is a technology that powers SDL’s content management and language solutions, and helps to process, understand and generate content, by finding patterns and connections within content across languages. By applying AI to content, we can understand its structure, language and intent which enables us to automate and scale content processes such as translation. By combining Linguistic AI with post-editing capabilities translations can processed faster at less cost while maintaining quality. It’s technologies like these that companies should explore if they want to address advanced content challenges, and take advantage of the opportunities of the vibrant, exciting intelligent translation era.

If you'd like to learn how SDL can help solve your MDR translation challenges, click here to learn more.

Eric Jacobs

VP, Life Sciences Strategy

With 15 years of direct Life Sciences globalization experience, Eric Jacobs has assisted many leading companies in the Pharmaceutical and Biotechnology sectors in successfully creating and implementing strategic initiatives across a wide spectrum of their ecosystem globally, from IMP labeling to regulatory filings. In his current role, Eric is responsible for outlining and managing the execution of SDL’s Life Sciences strategy globally and for driving vertical innovations across the entire portfolio of products and services.

All from Eric Jacobs