The Acolad blog

How to leverage machine translation for your medical translations

Written by Barbara Peralta | Jan 17, 2022 8:30:00 AM

Even with the advancements in machine translation, there are still a few industries where MT hasn’t caught on as much as it has elsewhere. It’s no surprise that with the sensitive nature of its content and the high-quality requirements in medical translations, life-sciences is one such industry.

However, the advent of neural machine translation has perhaps made that cautious approach obsolete. There are many cases in the medical industry where use of MT can add enormous value. Not only can it be deployed to reduce cost and turnaround times, but can also be used to expedite content analysis and address quality.

Practical applications of machine translation

Here are some ways in which machine translation can be practically applied:

1. Large volume programs

An obvious application of machine translation is when large volumes of content need to be translated or updated in a short period of time. We are seeing a lot of this type of work as companies race to comply with the MDR/IVDR deadline.

At Acolad, we have leveraged machine translation in medical translations to process large volumes of content to update product labeling in order to meet MDR requirements. With one global medical device customer, we processed over 424,000 source words into 39 languages in record time. This is something that would not have been possible without the application of neural machine translation.

2. Post Market Surveillance

As requirements around post-market surveillance become more rigorous and demanding, medical companies need a solution for sifting through vast amounts of content in many languages. Machine translation can allow for swift medical translations of content coming from patient complaints, adverse events, and even social media, allowing our customers to quickly focus on content that needs further investigation.

3. Clinical Trials

In clinical trials, where the timelines for producing localized content are extremely tight, neural machine translation is already making waves. Not only are results impacting timelines, but processing clinical trial documents, such as Patient Informed Consent Forms, can result in time savings that range from 30%-50%.

4. Content audits/quality reviews

Many of our pharmaceutical and medical device clients have been managing labeling in a decentralized manner – allowing local/in-country affiliates to manage and localize their own product content. This can often result in a loss of quality, loss of consistency, and introduction of patient safety risk.

Machine translation is the perfect solution for conducting content audits in many languages that can then be normalized for AI assessment of that content for risk management. In one example, Acolad was able to process a huge volume of local labeling in a time frame never possible before.

Can machine translation really be used for all types of medical translations?

Of course! The reality is that all content is a candidate for the application of MT. Patient-facing content can safely take advantage of the technology with the right approach to ensure high quality and low risk. This process should at minimum consist of some of the following:

  • Proper assessment of content to plot out machine translation suitability, process, and integration with pre-existing translated content in translation memory systems to maintain consistency with previously approved content.
  • Pilot testing of content and language combinations to measure quality and savings.
  • Post-edit of neural machine translation output by qualified medical translators
  • Integration of other linguistic assets that have been developed and approved over time – such as glossaries/term bases, translation memories, style guides – so that this investment is leveraged and applied in conjunction with MT.
  • Application of AI technology to further train and customize your machine translation engines for higher quality output.
  • Application of AI technology in other quality check applications to either streamline the post-edit task or automate quality checks.

The application of neural machine translation in medical translations is not just something to be considered for the future but is a reality now and should be an integral part of any translation program. Got questions? Connect with our experts!