How to Use ChatGPT For Medical Interpreter Terminology

Medical interpreting practice is a crucial activity for both beginners and experienced interpreters to develop the hard skills needed to perform their job successfully. A challenging aspect of any medical interpretation dialogue is researching terminology before, during, and after the interpreting exercise. Although researching medical terminology can be overwhelming for a beginner interpreter, technology can provide us with tools to make it a bit easier, and a great one is ChatGPT! 

ChatGPT in a nutshell…

You may have heard about ChatGPT on social media and may be left wondering, what is this newfangled ChatGPT, and why is everyone so polarized about it all of a sudden?! Some of your colleagues say that ChatGPT is an amazing program that will change our future, while others just see it as a run-of-the-mill kind of chatbot we should stay away from. But what is ChatGPT really?

ChatGPT is a language model developed by OpenAI that uses machine learning to generate human-like responses to text inputs. Essentially, it’s a chatbot that can converse with you in natural language. However, unlike other chatbots that rely on pre-written scripts or fixed responses, ChatGPT can generate unique responses based on the context of the conversation.

The idea of chatbots is not new. In fact, they’re one of the most popular types of artificial intelligence! For example, many companies use chatbots to provide customer support, answer questions, and help customers troubleshoot issues. What makes ChatGPT different from these typical chatbots is its advanced AI architecture called ‘’Generative Pre-trained Transformer (GPT)’’ to create more natural, human-like responses and engage more effectively in conversation. 

Zoom-s Virtual Chatbot
ChatGPT’s Desktop Interface

How can ChatGPT help me with my medical terminology as a medical interpreter? 

You may already know that searching for the most appropriate translation of a medical term within a specific context is an important but often time-consuming task. Think about the amount of time that you spend browsing different websites, dictionaries, and cross-referencing glossaries. All of those minutes can easily add up to the total time invested in researching just one term! 

One of the advantages of ChatGPT is that it has been trained on a vast corpus of textual data, including medical literature, journals, and research papers curated from sources on the internet, which enables it to answer medical-related questions within seconds.  While ChatGPT can’t completely replace the practical role of human-created sources like our specialized dictionaries and glossaries, it can certainly make our research process more efficient by providing us with helpful pointers on common terminology issues. 

Refreshing medical terminology 

Imagine that you’re in one of our live interpreting practice sessions with other InterpreMed members. While you and your peers are working on interpreting one of our original consecutive practice scripts, you encounter two frequently confused terms: sprain and strain. You have the right answer on the tip of your tongue but can’t really remember it. So, you decide to do a quick ChatGPT search to refresh your knowledge. ChatGPT will very articulately explain that a sprain is an injury to a ligament, while a strain is an injury to a muscle or tendon, as you can see in the detailed answer below:

A sprain and a strain are both common injuries that can occur in muscles and joints, but they are different in terms of which part of the body is affected.

A sprain refers to an injury to a ligament, which is a tough, fibrous band of tissue that connects two bones together. Sprains usually occur when a joint is twisted or stretched beyond its normal range of motion, causing damage to the ligaments. Common locations for sprains include the ankle, wrist, and knee. Symptoms of a sprain may include pain, swelling, bruising, and difficulty moving the affected joint.

On the other hand, a strain refers to an injury to a muscle or tendon, which is the tissue that connects muscles to bones. Strains usually occur when a muscle or tendon is stretched or torn as a result of overuse or sudden force. Common locations for strains include the back, hamstring, and shoulder. Symptoms of a strain may include pain, stiffness, weakness, and muscle spasms.

Not bad at all! If you’re feeling curious, you can also ask ChatGPT simple follow-up questions, such as typical treatments and diagnostic tests for sprains and strains, and it will give you an answer drawn from the medical information available in its database which, as of today, doesn’t contain information past September 2021.

However, ChatGPT can’t replace our medical libraries like MSDManuals, MayoClinic, or MedlinePlus, nor the sound knowledge provided by a good medical doctor. It might be helpful to think of ChatGPT as an AI-powered search assistant. Beware that ChatGPT is especially skilled at providing incorrect answers in the most convincing and confident manner! Just like anything we search on Google, we can’t take the information we see at face value! So, a healthy dose of skepticism and critical thinking will go a long way in seizing ChatGPT’s advantages!

Looking up different registers 

Now imagine you’re practicing simultaneous interpretation with one of our recorded monologues. You come across a colloquial term for which you only know the formal translation. Although you interpreted the term’s meaning correctly, you still feel like there’s a better way to express it using the same register as the speaker, or maybe you’re just curious. ChatGPT can help you solve this linguistic puzzle! For example, if I wanted to find idiomatic ways of saying ‘’sprain’’ or ‘’strain’’ I can use the following prompt:

‘’Give me 10 idiomatic phrases for ‘sprain’ and ‘strain’ in the U.S.’’ 

And it will retrieve:

To pull a muscle

Tear a muscle

Overexert oneself

Overdo it

Push oneself too hard

Strain a muscle/tendon

Damage a muscle/tendon

Overwork a muscle

Tax a muscle

Put too much pressure on a muscle

Reviewing this variety of phrases and related words can be incredibly helpful in expanding your vocabulary within multiple registers: from the colloquial and everyday speech that you typically hear patients use to the technical and formal lingo used by healthcare providers. All in one quick itemized list! Keep in mind that ChatGPT doesn’t take into account important context clues such as purpose and location by default, so it’s best practice to write very explicitly in your prompts. For example, ‘’Give 10 most commonly used slang in U.S. English that means ‘to wake up early’. You’re a rural and witty 17-year-old from the South of the United States in 2020’’. You can have fun and try your best to read the following phrases using your best Southern accent! 

Rise and grind: “Time to rise and grind, y’all! Gotta tackle the day head-on.”

Crack of dawn: “I always wake up at the crack of dawn to get a head start on things.”

Up with the chickens: “I’m always up with the chickens, ready to start my day bright and early.”

Rise and shine: “Rise and shine, folks! It’s a brand new day full of possibilities.”

Crack of sparrows fart: “I woke up at the crack of sparrows fart to make the most of the morning.”

Early bird gets the worm: “I believe in the saying, ‘The early bird gets the worm,’ so I always wake up early.”

Up at the butt-crack of dawn: “I dragged myself out of bed at the butt-crack of dawn, bleary-eyed but ready to go.”

Up before the roosters: “I’m always up before the roosters, raring to start my day.”

Rise and shine like a rooster on a hot plate: “I woke up in a hurry, rise and shine like a rooster on a hot plate!”

Up and atom: “Time to be up and atom, folks! Let’s get this day started.”

These are just a few examples of Southern slang; it’s by no means an exhaustive list, as individuals may have their own slang preferences, but it’s certainly a great start! Note how specific I was in my prompt, including age (17 years old), year (2020), English variation (rural U.S. English), and frequency (most commonly used). These are some contextual elements you want to specify to ChatGPT to consider in its response. Keep in mind that ChatGPT’ has not been trained on data past September 2021, so it may not help with very recent terms. 

Creating glossaries 

After you’ve finished your medical interpreting practice session, you and your peers are ready to wrap up the meeting! But before you set a date and time to work on the next medical interpreting dialogue in our interpreting study guide, you want to save a list of medical terms that were especially challenging for you. If you like keeping digital glossaries, ChatGPT can be very helpful for you! 

Imagine that the following terms were the ones you struggled with the most during the interpreting exercise: herniated disk, heating pad, corticosteroid injection, no walk in the park, anesthesiologist, and formwork carpenter; and you want to practice with these specific terms in the future. 

You can prompt ChatGPT to create a table using your specific parameters! For example, “Create a table with the following words herniated disk, heating pad, corticosteroid injection, no walk in the park, anesthesiologist, and formwork carpenter. The first column is a term from the list and name it ‘English term’, for the second column, leave the cells blank, and title that column ‘’Spanish translation’’, the third column is titled ‘’Definition,’’ and you’ll add a brief definition of the term in English.’’

You can then easily copy-paste this list to your preferred terminology management program or system. I, for one, like to use Google spreadsheets, but you’re free to use the method or program you’re most comfortable with! 

Respecting confidentiality as a medical interpreter

As we use more and more digital tools, it’s important to be mindful of how our data is used online. This is usually forgotten by the appeal of free digital tools automating so many of our work tasks, but there are still important confidentiality issues to keep in mind. Per ChatGPT’s privacy policy, it may use the content of the messages you send for a variety of purposes, including disclosing it to third-party vendors. For this reason, you must refrain from sharing private information about the clients or patients you may work with to comply with both the ethical principle of confidentiality and HIPAA rules in the U.S. 

Final Thoughts

A phrase has grown popular amidst the fear many linguists feel about AI taking our jobs: AI will not replace you; someone who uses AI will! 

You can sign up for ChatGPT here and pair this powerful tool with our original interpreting practice materials! Become a member today to make the most of your live interpreting practice sessions and self-paced interpreting exercises in the consecutive, simultaneous, and sight translation modes for medical and educational interpreting

If you enjoyed this blog post, don’t forget to subscribe to our newsletter and receive email updates when our next blog post is published! 

Leave Comment

Your email address will not be published. Required fields are marked *

You cannot copy the content of this page!

Log In