It turns out that a little bit of math can both improve the results of mammography and make expert radiological analysis available to more people in remote and poor areas of the world.
Researchers have known for a few years now that applying a mathematical transformation method known as "wavelets" to radiological images can improve the ability of doctors to detect cancer. But Bradley Lucier's team of mathematicians at Purdue has taken the process to a new level -- by using the wavelets method to compress mammogram images by 98%, not only can radiologists still detect cancer better than they can with unmodified images, the mammograms become small enough to send easily over the dial-up computer networks common in poorer parts of the world. The work will appear in the next edition of Radiology.
"Any technique that improves the performance of radiologists is helpful, but this also means that mammograms can be taken in remote places that are underserved by the medical community," said Lucier, who is a professor of mathematics and computer science in Purdue's College of Science. "The mammograms can then be sent electronically to radiologists, who can read the digitized versions knowing they will do at least as well as the original mammograms." [...]
Lucier is optimistic that the technique might be applied to other forms of telemedicine as well, if certain modifications are made.
"There are many forms of medical diagnosis that require an image to be read by a specialist," he said. "If image compression is applied to other diagnostic situations, you won't actually have to have that specialist on hand if you can get the equipment to the patient. But this is proof in principle that file compression, if done properly, can confer advantages to both patient and doctor."
A single uncompressed mammogram can run up to 50 megabytes in size, and diagnosis typically requires four different images. The wavelet process cuts the file size down to approximately one megabyte per image, well within the capabilities of most dial-up or even cell-phone Internet connections. Although other researchers have demonstrated that the use of wavelets can improve radiological diagnosis, Lucier's group managed to shrink the image files far more than ever before, using an algorithm Lucier himself created a decade earlier.
This is one of those breakthroughs that, on the surface, appears quite technical and obscure, but when boiled down to the basics, actually has direct benefits for many people. It's easier to mammographic and other cancer detection gear distributed to poorer parts of the world than it is to get expert radiologists; this discovery means that telemedicine can become a viable tool for cancer detection, even in areas out of the reach of high speed Internet links.
The real issue lies at how many hospitals in the poor countries can afford the mammography machine and the high resolution scanners. Medical equipment is still a high margin business and the price remains stiff (that's why GE dumped the electric applicances and moved into it in the 90's). If Dell is wise enough to enter this industry, the price of a low end medical imaging machine might plough. With the threat of $100 laptop, Dell should probably exploit the medical equipment market, just as it moved into printer, pocket PC and LCD.
Dell is famous for its 2% R&D budget out of its revenue but medical equipment requires much more R&D resource than that. See if Kevin Rollins is visionary enough to take this move.
Actually, having just been in Guyana, much of the radiology is outsourced. There typically are centers where mamography is done, and the files are sent to radiologists via the internet. The smaller file size is a big deal.
An algorithm that cuts something down to size is exponentially more valuable than more powerful hardware. ;-)