Method facilitates creation of computer models of the spine

Set of equations developed by Brazilian researcher allows for estimation of vertebrae measurements and could guide medical treatment.

By Karina Toledo

Agência FAPESP – A set of equations developed by a Brazilian researcher during her doctorate at Ulm University in Germany has made creating computer models of the human spine easier to use for testing new implants and simulating surgery.

The thesis presented by Maria Elizete Kunkel has resulted in the recently released book Biomechanical models of the human thoracic and lumbar spine: A statistical approach for prediction of anatomical parameters from radiographic images, published in English by Germany’s Südwestdeutscher Verlag für Hochschulschriften.

“The biomedical model allows for simulation of all the spine’s movements on the computer. This helps the doctor to decide, for example, which type of implant is best suited for a patient that needs corrective surgery. The model could be useful in the treatment of fractures, herniated disks and other postural disorders, such as scoliosis, lordosis or cyphosis,” says Kunkel.

This model, however, must be personalized according to the vertebral dimensions of each of patient. “Normally, the patient is submitted to computerized tomography and afterward the image is reconstructed using computer resources. However, this is expensive, time-consuming and exposes the patient to high doses of radiation,” she said.

The original idea for her doctorate, says Kunkel, was to seek the quickest and most financially accessible manner to obtain these models. “However, I had a lot of difficulty feeding the program that was being developed with the data. I needed to have patient measurements,” she said.

The researcher then contacted a group of scientists that studied the vertebrae of 12 human cadavers without any spinal problems. “They measured the 17 vertebrae of each person in detail. There were 40 dimensions measured in each vertebrae studied,” explains Kunkel.

The work resulted in a table with the average vertebrae sizes, which was the basis of Kunkel’s studies. After detailed statistical analysis, she realized that one of the measurements – the height of the vertebrae – had a good relationship with the other 39 dimensions.

Based on this value, the researcher used regression analysis to develop 50 equations that made it possible to estimate the value of the other measurements. 

“The major advantage of the method was that the height of the vertebrae can be obtained with a simple X-ray exam, which is very accessible. There is clearly a margin of error, but we conducted tests and verified that this margin of error was lower than when all the dimensions were directly measured in the patient,” she affirmed.
 
The study resulted in three articles published in the Journal of Anatomy. “My adviser recommended that I dedicate my time to creating these equations and this part of the study ended up becoming my doctorate. The creation of a biomechanical model was left for another researcher to finish,” says Kunkel.

The thesis received top marks, and Kunkel graduated Magna Cum Laude from the university. The thesis also caught the interest of editor Südwestdeutscher Verlag für Hochschulschriften.

Holding an undergraduate degree in physics from Universidade Federal do Ceará, Kunkel completed her master’s at Universidade de São Paulo in São Carlos (USP- São Carlos), where she developed the project “Comparative evaluation of bone ultrasonometry measurements with mechanical compression essays to estimate fracture risk,” which was funded by FAPESP. She also completed a Scientific Journalism course at Universidade Estadual de Campinas and undertook a project with FAPESP funding.

• Biomechanical models of the human thoracic and lumbar spine: A statistical approach for prediction of anatomical parameters from radiographic images
  Author: Maria Elizete Kunkel
  Release: April 2012
  Price: US$ 77.00
  Pages: 80
  Available at Amazon.com.