MACHINE TRANSLATION
AND POST-EDITING (MTPE)

Discover the precision and innovation of individual
AI solutions.

icon
background image

Revolutionizing medical translation:Machine translation in the health care sector

Machine translation revolutionizes cross-lingual communication in medicine by automating text translation. It analyzes the input text using algorithms to produce translations of varying complexity, from the simple to the more nuanced. Accurately capturing medical context and cultural nuances remains challenging and requires human post-processing. However, thanks to continuing advances in machine learning and data availability, machine translation is poised to significantly impact medical communications worldwide.

Custom AI solutions at mt-g

We at mt-g support global medical communication with machine translation and rely on modern AI solutions that ensure accurate, effective, and safe communications for health care and the life sciences without compromises!

 

Machine Translation & Post-Editing in accordance with ISO 18587

If the source text is suitable for machine translation, then the desired quality level for the post-editing step is ascertained. As can be seen in the diagram above, either light post-editing or full post-editing is applied depending on the needs of the project.


If the source text is not suitable for machine translation, a human translation will be proposed.
 

Learn more about translation

The benefits of MTPE

Cost and time savings
by delivering the translation quickly
without affecting the cost.

Scalability
due to the ability to
handle large volumes of
text translation in a
very short space of time.
This allows companies to scale their projects
efficiently.

Quality
by providing the best possible combination of
artificial and human intelligence.

Data protection
of personal data in
accordance with
the GDPR.

mt-g: YOUR RELIABLE PARTNER

for machine translation and post-editing in medicine

Our professional posteditors are specialist medical translators. Thanks to our careful post-processing, you will get a result that is the equivalent of human translation.

Our services include both pure post-editing of already machine-translated texts and a comprehensive MTPE package of services. Together we can devise a tailor-made solution for your project. This will let you get the most out of machine translation while maintaining the same quality that you would expect in a human translation.
 

FAQ

Yes to the last question. The customer may order these services either individually or as a combined package.

If the customer wishes to order just post-editing because they did the machine translation themselves or obtained it elsewhere, then we would need to review the result in order to estimate the expected effort for post-editing and to draft a corresponding proposal. If the provided machine translation is unusable, one of our project managers will contact the customer and discuss possible solutions.

Customers can choose between two options:

 

OPTION

ADVANTAGES

DISADVANTAGES

APPROACH

BASIC

Fast and ready to be deployed, wide contextual coverage, and cheap

No training, no specialization, low adaptability, and standard quality

General context with sparse terminology density, limited data resources, and quick turnaround

CUSTOM

Can be trained (using translation memories, stock, language corpora), customizable to industry, exact translations

Dependent on data availability and training process, and a bit long and complicated at the beginning

Ongoing translation needs, high demands for translation accuracy


Customers can also choose between two options for how to allow contractual partners to process data:

OPTION

APPROACH

Cloud

  • Hosted on cloud servers, managed by a third party
  • Scalability and flexibility
  • Most cost-effective option

On-premises

  • Hosted locally using the company's own infrastructure
  • Control of data and infrastructure, provision of own security measures
  • Upfront investments in hardware, software licenses, and establishment of infrastructure, meaning it is more cost-intensive

 

 

There is no single answer to the question. Any savings from MTPE depend on a variety of factors. The costs are significantly influenced by customer decisions (which AI solution, which post-editing level), the effort required to achieve the desired quality (data preparation, frequency of training, creation and maintenance of glossaries), the chosen infrastructure (cloud or on-premises), the extent of the existing translation memories (the larger the translation memory, the lower the cost advantage of machine translation), and, of course, the quality of the machine translation results (requiring more/less complex post-editing).

In addition, the cost advantages should be considered over the long term: With on-premises solutions, for example, the initial effort and investment are slightly higher at the start, but as the projects progress this initial outlay can really pay off. 

 

First and foremost, we work with some of the best machine translation providers in the language technology industry. Nothing is returned to the customer without human review and post-editing to ISO 18587 standards by our language experts.

Our dedicated machine translation team continuously tracks output using both automated quality assessment metrics and review by human language experts. By bringing people and cutting-edge technology together, we ensure that we are able to work efficiently while delivering high-quality translations.
 

Yes. We can issue letters of confirmation for machine-generated and post-edited translations. If post-editing has been applied, then the confirmation certifies compliance with ISO 18587 and clearly indicates the post-editing level that was applied.

Personal data in medical texts is processed by mt-g in accordance with the provisions of the General Data Protection Regulation (GDPR).
Data is stored in secure and controlled environments with strict measures to protect against unauthorized access or breaches (using on-premises solutions).
 

Machine translation struggles to resolve ambiguity and can produce inaccurate or misleading translations, especially for source texts with words that can mean different things depending on context (such as marketing content). Furthermore, the machine has no cultural awareness about how to accurately translate context-specific references. Machine translation can also show its limits when it tries to make sense of technical factors, such as when transcribing timestamps and or processing software strings with character limits.

Contact us

Schedule a personal consultation or submit a request for a quote for a specific project.

You can learn more about how personal data is used in our privacy policy.