back to top
How to Use AI to make Stunning Medical Illustrations. AI text surrounded by medical graphics

How to Use AI to make Stunning Medical Illustrations

Stephan Kuslich
October 24, 2022
View Table of Contents

AI Art, Medical Images Generated from Text, Midjourney vs. Stable Diffusion vs. Dall-E 2

Learn how to use artificial intelligence to automate the creation of medical illustrations, anatomically correct images of anatomy, pathology and surgical treatments.

Computers have been used to assist in the creation of medical illustrations for years. Automation tools like automatic tweening can make medical animations much faster to produce. Even basic features like UNDO can save artists a lot of time and effort. However, these tools still require a skilled artist to use them properly to create anything of value.

We’ve all been told that AI will eventually take all our jobs. However, most of us have considered our careers safe from this threat, as they require creative thinking that computers cannot yet provide. I’ve been working in the field of computer graphics for over 3 decades and have seen few examples of software being able to create anything remotely creative without human intervention. That is until just very recently.

Nearly every phone and free photo app features some form of Auto-Retouching that can remove blemishes and balance even the worst exposure problems. Fans of Adobe Photoshop have collectively saved millions of hours of manual image tracing using the notorious Magic Wand Tool to remove unwanted elements from photos in a single click. Power users can direct Photoshop to greater feats of automation using the Actions and Batch Automator to process thousands of images as fast as their computers can crunch the pixels. 2D and 3D animation software has made the painstaking process of drawing thousands of frames of images just to produce a couple of minutes to a process of setting a few key frames and letting the computer fill in the stuff between.

I’ve now tested several new AI image generators that create completely original works of art from words, and the results have been surprising. Even though it’s not perfect, it is much more functional than I thought it would ever be in my lifetime. Depending on the words you use to describe the image, you might get exactly what you’d expect, or you could be completely confused by the generator’s interpretation.

The results of AI art generators can be surprising.

What would you expect when you ask for, “a map of human anatomy?

I often get art back from medical illustrators that doesn’t capture what I was hoping to see. Both human illustrators and AI might need a few tries to get it right, but the difference is that AI can deliver a completely different set of color images in just a minute or two, while human artists might need a day or two to make revisions. Also, AI image generators are so much cheaper than human artists that they might as well be considered free. This low cost and high speed allows directors to try things with AI that they’d never even consider attempting with a human illustrator.

Create a significant volume of concept art in very little time

With nothing more that a few carefully chosen words, AI was able to create this collection of orthopedic pathologies in 30 minutes.Even the most experienced illustrators can become frustrated when they see no end to the tweaks and alterations, especially when they can tell that a director is just trying out different concepts.. I’ve watched veteran artists revert into colicky babies when asked to make revisions to details they don’t fully understand. AI doesn’t care about any of that. For directors who want to visualize produce a wide range of client samples, AI won’t stop until the director stops

Arthroscopic view of anterior cruciate ligament
osteoblast and osteoclast 3D rendered looking microscopic image
osteoporosis physiology
Spineology OptLIF spinal implant

The three leading AI image generators are Midjourney, Stable Diffusion, and Dall-E

Below is an example of the output of each rendering various art of a micro-biology image.

The three leading AI image generators are Midjourney, Stable Diffusion, and Dall-EBelow is an example of the output of each rendering various art of a micro-biology image. Comparing 3 Different AI Image Generators to Create Medical Illustrations
Prompt Used Midjourney AI Stable Diffusion AI Dall-E AI
Even the Best AI Image Generators have Limitations
Hands and fingers are particularly challenging for AI Image Generators to create.
AI Image generators often attempt to create esoteric images over literal image

Tying to get AI to create an image that depicts CPR is more challenging than you might imagine.

We’ve found Midjourney to be the most versatile of all the leading AI image generators. Learn more about using Midjourney to make medical graphics.

Blog with Author Section

AI Image Generators for Medical Illustrations

Research and Experimentations with Midjourney, Stable Diffusion, and Dall-E

Stephan Kuslich painted portrait, company founder

Stephan Kuslich

Company Founder and CSO

View Bibliography

Similar Blog Content

Browse More Categories

Talk To Us

Ghost Productions is an award-winning medical media production company that can help you with custom medical animation and marketing services. We offer a wide array of digital representation services that can help anyone in the medical field improve their marketing, support patient communication efforts, or even streamline staff training. Our team is highly trained and has a great understanding of biomedical processes, so they can accurately represent your device or product in a 3D medical animation, or other types of illustrations you may be interested in. Contact Ghost Productions now and tell us more about how we can help you train your surgeons, sell more medical devices, or explain your pharmaceutical product.