One lament I’ve been hearing pretty consistently from medical communication colleagues is how technical discussions about artificial intelligence (AI) can be, with terms that go undefined and examples that are not relevant to our field or industries.
To address these concerns for my colleagues in medical communication (freelance or in-house), I’m adding a glossary of AI terms for medical writers here on the DCC Cyber site.
Today I’m sharing the first 10 entries, the “Top 10” entries that really are salient to anyone, but with examples in the medical context.
The main difference between this glossary and content you will find at other resources (even ones for writers) is that I arrange the terms in an order of importance for medical writers and editors and write the related entries to be consistent with one another. For example, algorithm, training, model, and application are related terms that are part of a process, so they are written together using the same example.
In the interest of time, I’m not sure I’ll be able to provide examples for the entire list as I go, but I’ve added them to these first 10 to get things started.
So, without further ado, the Top 10 AI Terms Every Medical Communicator Should Know.
1. Artificial Intelligence (AI)
The simulation of human intelligence in machines designed to perform tasks like learning, problem-solving, and decision-making.
Example: AI-powered systems that can detect diabetic retinopathy from eye scans.
2. Machine Learning (ML)
A subset of AI that allows computers to learn from data patterns and improve performance over time without being explicitly programmed.
Example: ML algorithms that predict patient readmission risks based on historical hospital data.
3. Deep Learning
A type of machine learning based on artificial neural networks, particularly effective in image and speech recognition.
Example: Deep learning models that analyze chest X-rays to detect signs of pneumonia.
4. Neural Network
A computing system inspired by the human brain’s structure, composed of layers of nodes (neurons) that process information.
Example: Neural networks used to predict the likelihood of a patient developing sepsis based on real-time vital signs and lab results.
5. Natural Language Processing (NLP)
The branch of AI focused on the interaction between computers and human language, enabling machines to understand, interpret, and generate human language.
Example: NLP systems that automatically extract relevant information from clinical notes in unstructured (eg, free-form text fields) electronic health record (EHR) fields so the data can be used in analysis.
6. Large Language Model (LLM)
An advanced AI system trained on vast amounts of text data that can perform tasks such as answering questions, generating text, translating languages, and summarizing content.
Example: Using an LLM to summarize a complex research paper to generate an accurate, simplified version of the paper’s findings for a general audience.
7. Algorithm
A set of rules or instructions for a computer to perform specific tasks, like performing a calculation or learning patterns from data.
Example: The set of steps or rules that guide how a system processes skin lesion images. It could be a machine learning algorithm that knows how to analyze image features (eg, color, texture, shape) to classify the lesion as benign or malignant.
8. Training
AI training is the process of teaching algorithms to make decisions by exposing them to large amounts of data. This allows the AI to recognize patterns and learn from examples.
Example: Feeding a computer program thousands of labeled images of benign and malignant lesions to train the AI to recognize patterns in the data.
9. Model
In AI, a model is created by training an algorithm to learn patterns and relationships from the input data. Models are the core of AI systems and can be trained for specific tasks, such as diagnosing diseases or predicting treatment outcomes.
Example: After training an algorithm with thousands of labeled images of benign and malignant lesions, the result is a model. This model has learned to recognize patterns in the data and can now predict from its image whether a new lesion is likely to be benign or malignant.
10. Application
An application is a software tool or program (“app”) that uses models in real-world practice.
Example: A health care provider can upload a patient’s skin lesion image into an app on their tablet, and the AI provides a prediction (eg, “high likelihood of melanoma”) to assist with diagnosis and treatment decisions.