Machine Translation of Medical Content at Scale

If you need to translate a growing amount of medical content into multiple languages, consider deploying machine translation (MT) and post-editing.
As patients increasingly turn to the Internet to do a preliminary search on their condition or on behalf of a loved one, it has fueled a demand for more and more medical information to be made available online. Patients, as well as practitioners using web-based medical resources, expect this information to be available in their native language. 

What Is Machine Translation?

Machine translation (MT) is the use of software to translate text from one language to another. It is a versatile localization instrument that can help:

• lower the cost of a localization project
• increase productivity 
• decrease project timelines

What is Post-Editing? 

Post-editing is the correction of machine-translated text by experienced and qualified human medical translators to ensure that it meets the agreed-upon level of publishable quality requested by the client. 

What Types of Medical Content Can Be Machine Translated?

Most types of medical content are suitable for machine translation and human post-editing, including:

• Educational and training content: research papers, educational websites, and training videos
• Clinical trials: regulatory documents and patient records
• Medical devices and apps: user guides and instructions for use
• Biosciences: studies, research reports, and papers
• Web: medical help sites and FAQs
• Medicines: inserts and descriptions

How to Test if Medical Content is Suitable for Machine Translation?

If you are considering machine translation and human post-editing to translate your medical content, it is essential to first determine whether your medical content is suitable for MT. 

At SimulTrans, we can:

• Test a small sample of your medical content to ensure it is suitable for MT
• Assess how much effort it would take to have a professional medical post-editor correct the machine-translated content to meet your desired publishable quality
• Supply you with a Machine Translation Suitability Report outlining our findings and advising you on how to proceed depending on the volume of content, the type of materials, and the target language(s)
• Calculate the savings of deploying machine translation and human post-editing compared with traditional human translation and review

How to Run a Pilot for a Machine Translation Project 

Once you have this comprehensive Machine Translation Suitability Report from SimulTrans, you will know if your medical content lends itself to machine translation + post-editing, which languages are best suited, and which languages would require considerable post-editing. Then you could:

1. Run a small machine translation + post-editing pilot project in 1-2 languages
2. Send the translated content to the doctors in your organization for review
3. Conduct a quantitative survey to obtain practical and precise information
4. Compile valuable insights and data from the feedback to share with management

How to Make a Case for Deploying Machine Translation

The feedback from the pilot test from doctors, together with numerical data from the survey results, should help you make a case for using machine translation and human post-editing at your company.

Next, you should research the level of localization required:

• Estimate the total volume of words
• Create a priority list of content 
• Select the best languages to target 
• Decide what types of content to machine translate
• Figure out how to handle content that is not suitable for machine translation
• Draft a budget and get approval from key stakeholders

How to Deploy Machine Translation at Scale

Once you have the go-ahead to implement machine translation + post-editing at scale for your organization, SimulTrans can customize the machine translation program by:

• Selecting the best Neural Machine Translation(NMT) engine for your project
• Reviewing and preparing all content for machine translation
• Creating glossaries, reviewing existing translation memories, importing style guides, and merging term bases
• Applying ISO 18587 standards to select and train a team of human post-editors in your subject matter
• Creating translation workflows to handle minor updates and full rewrites 
• Automating quality checks to help with post-edit tasks
• Fine-tuning workflows to improve the localization process

Ready to implement AI translation and post-editing as part of your medical translation workflow? Start by requesting a free machine translation suitability report.