Warning: the following videos are graphic!

Digital image morphing of a bite from a brown recluse spider, progressing from a minor wound to an extensive necrosis of the skin and underlying soft tissues.

Successful healing of frostbite on the fingers.

You’ve probably seen digital image morphing before, either in commercials or in the movies. Hollywood has used digital image morphing for years now to turn humans into animals, humans into other humans, or to simulate aging. Now a paper by Keith Penska, Les Folio, and Rolf Bunger, entitled ‘Medical Applications of Digital Image Morphing’ in the Journal of Digital Imaging (2007), introduces digital image morphing to medicine.

Digital image morphing is a process by which one image transforms into another image through a series of generated intermediate steps. This is performed through a program that generates a smoothing algorithm that makes the final video seem to blend slowly from the initial image into the final one.

All you need to create an image morph is readily available software (just search for image morphing software on Google) and an average computer or laptop. To ensure a good image morph you do need to make sure images are of equal size and properly registered with each other. This will create a smooth transition. Also, make sure the backgrounds are the same. You can tell in the above videos that the images weren’t completely registered.

Digital image morphing has potential medical applications, especially for primary care physicians and their patients. It provides a quick and cost-effective way to show patients how their treatment has progressed over a period of time, or how their disease has progressed over time. And this doesn’t just apply to diseases or injuries on the surface of the skin. Image morphing can be performed with x-ray’s and even with MRI and CT scans. Plastic surgeons can use image morphing as a sort of digital portfolio to show before and after transitions. Image morphing provides a more dramatic effect than viewing two separate images of a before and after.

However, there are limitations. According to the journal article mentioned above, “the smoothing process is synthetic and based on information from each of the images control points to reflect the blending of the images as the movie is created.” Basically this is saying the images in between the original images are not real, they are only an approximation. Obviously, the more images you have in the sequence, the more accurate the representation.

I’d like to hear what you think. Is this a good idea? Would doctors use this technology to educate their patients?