Mathematical Analysis Shows Early Bone Deterioration Not Detected Byosteoporosis Tests

BUFFALO, N.Y. -- Fractal geometry may hold the key to earlier detection of osteoporosis, according to research conducted at the University at Buffalo.

It may also provide a better understanding of how microgravity affects bone structure.

UB bioengineers will describe a new computational tool they developed and tested that characterizes bone deterioration more accurately than methods currently used at a conference on Physiology and Function from Multidimensional Images to be held Feb. 26 in San Diego.

The technique is based on evidence that bones are composed of fractal structures. Fractals have a property called self-similarity, which means they retain the same properties regardless of the angle, scale or resolution from which they are viewed.

"Bones are composed of two types of tissue -- solid cortical tissue and interconnecting honeycomb structures, called trabecular tissue," explained Raj S. Acharya, Ph.D., associate professor of electrical and computer engineering at UB, and principal investigator. "It is these honeycomb structures that are similar to fractals."

Acharya explained that clinicians who look at X-ray images of bones have only been able to track deterioration in bone mass; deterioration in the honeycomb structure has been too subtle to see.

Now, Acharya and his colleagues have found a way to examine the fragile interconnections in this tissue and to determine the level of deterioration.

Using new algorithms they devised, they found that the technique lets them home in on each pixel, or piece of visual information in an X-ray or magnetic-resonance image, and assign to it a specific value called a fractal dimension.

"With fractal dimensions, we can compute structural information," Acharya said. "Together, mass and structural information give a much better indicator of bone condition."

Working with animal data obtained from Merck Corp., Acharya and colleagues from the UB School of Dental Medicine have used fractals to compare the trabecular tissue of the bones of normal rats with those of rats with osteoporosis induced through estrogen depletion or the immobilization of limbs.

"In both cases, when we computed the fractal dimension in these data, we found there was a significant difference between the values for normal bones and the values for bones with osteoporosis," he said.

Once bone mass begins to deteriorate, it usually means osteoporosis has progressed to an advanced stage.

But in the early stages of the disease, it is the honeycomb structure that provides the most important information.

"In these stages, the honeycomb structure might actually be broken, which would weaken the bone, but because the mass might still be the same, a bone density test would not show anything abnormal," he said.

"The advantage of our approach is that if you can pick up signs of disintegration right at the beginning before the bone starts massive deterioration, then you can start therapy," he said. "And if you catch it at an early stage, there's a good chance that a deeper osteoporosis will not develop."

In ongoing collaborations with researchers at the NASA Johnson Space Center and Baylor College of Medicine, Acharya is also analyzing how bones deteriorate due to microgravity conditions in space. The work involves computing the fractal dimension of magnetic-resonance images of the bones of human subjects who have undergone prolonged bedrest. The researchers are quantifying the deterioration in the trabecular bone structures, as well as how exercise and drug therapy affect that deterioration.