Kinect could help medics estimate radiation dose from CT scanner

Researchers at the Hospital of the University of Pennsylvania (www.pennmedicine.org/hup/) are investigating whether a system based around the Microsoft Kinect could be used to help predict the precise dose of ionizing radiation that patients require during a CT scan by providing an accurate estimate of whole-body volume.

Although the American Association of Physicists in Medicine has developed a means to estimate the size of a patient for the purpose of calculating such radiation doses, this is not a volumetric measurement and can lead to an underestimation of the dose in smaller patients and overestimation in larger patients.

To address these limitations, Dr. Tessa Cook and her colleagues used a Microsoft Kinect system to capture depth information from a number of patients. An algorithm developed by Radimetrics (www.radimetrics.com) then created a depth map and a skeletal outline of each individual, both of which were used to estimate the whole-body volume of an individual.

Using the system, the researchers evaluated the effect of the position of a number of individuals by estimating their whole-body volume when their arms were by their sides, extended above their heads, or crossed over their chests. The density of an individual was then calculated by dividing each individual's weight by their volume.

The estimation of the volume of the patients varied according to the position they took in front of the camera. When their arms were extended over their heads, their bodies appeared to occupy a larger volume, resulting in a lower density estimate. Similarly, when their arms were crossed over their bodies, they appeared to occupy a lower volume that increased the estimates.

To improve the accuracy and consistency of volume estimates across different positions, Dr. Cook says that perpendicular depth maps may be necessary, obtained either with two Kinect cameras positioned at a 90° angle to one another, or by recording depth maps with the patient in consecutive frontal and lateral positions with respect to a single camera.

Further work is also needed to explore the differences introduced by patient height and age and to determine what additional modifications are necessary to accurately estimate volume in these individuals. More information can be found athttp://bit.ly/MCXSXE.