I'm trying to prototype a burn-surface-area determination application for our pediatric surgery unit, and can't seem to find a "real" body-scan that would be applicable. I can easily pull a skin mesh off of a MRI or CT dataset, but for the prototype I really only need something on the order of 256 x 256 x 512 voxels, which makes using something like the Visible Human complete overkill (as well as not being pediatric).
How much of the body do you need, i.e., is torso without full arms enough? A true whole body scan does not happen too often in standard clinical practice.
Also, assuming torso is enough, do you care what age? We have a range starting at about 2 years of age.
Rob
I work with the Office of the Medical Investigator here in Albuquerque (affiliated with the University of New Mexico) and we do whole body CT of all deceased subjects, including pediatric cases occasionally. We also do MRI on select cases. If you look me up on the web (UNM Dept. of Biochemistry) and send me your email, I can see if we can send you a de-identified whole body image. What age are you most interested in? (infant, toddler, 6 year old??)
Natalie
Many thanks!
Ideally we could use the whole body, and ideally a range of ages and weights/morphologies, though anything at all would be helpful and most appreciated.
The project is for a our pediatric burn unit. They're finding that the historically accepted methods of determining burn surface area (as a percentage of the body) in pediatric subjects, is becoming less and less valid, as children are becoming "better fed". This is leading to a host of problems starting from resuscitation, where accurate determination of the burn surface percentage, versus the body weight, is critical for determining the amount of fluids to provide IV, to surgery where accurately ordering or culturing skin for grafting is becoming more difficult, leading to either wasted graft material, or secondary surgeries when a graft turned out to be insufficient.
My group is trying to come up with a usefully quantitative way to predict skin surface-area from a limited subset of body-morphology measurements, but to do that, we need some basis for ground truth. My current best-guess for acquiring ground-truth is extracting the skin layer from MRI or CT data. There's probably something faster and cheaper that could be done to acquire a good complete data set once we prove that better quantitation is possible, but MRI and CT seemed like the most likely sources of data that might be already acquired for some other purpose.
Thanks again,
Will
That all makes good sense. I'll be meeting with Center for Forensic Imaging team members today and will find out if we can share de-identified images with other researchers. If so, I'll request whole body pediatric images representing a range of ages and body weights. Always feels better when we can do something to help the living!
Have a look at the Virtual Family: http://www.itis.ethz.ch/services/anatomical-models/overview/ and Virtual Classroom http://www.itis.ethz.ch/news-events/news/virtual-population/virtual-classroom/. You can get them for free (just pay for shipping of the DVD) and the models come as segmented 3D CAD models and a software that you can use for voxel voxel export at any resolution required (down to 0.5 mm³, but also low-res as you would prefer). I'm working with those models regularly (for MRI coil simulations) and think they're quite good.
You can download from following website
http://www.osirix-viewer.com/datasets/
Have a look into this paper, interesting methodology. might be of help for you. Author work in the NICU at Frankfurt University, Germany
Neonatology. 2011;100(3):260-4. doi: 10.1159/000327516. Epub 2011 Jun 22.
Three-dimensional body scanning: a new method to estimate body surface area in neonates.
Schloesser RL, Lauff M, Buxmann H, Veit K, Fischer D, Allendorf A.
Source
Department of Neonatology, Children's Hospital, Johann Wolfgang Goethe University, Frankfurt am Main, Germany. [email protected]
Abstract
BACKGROUND:
Body surface area (BSA) is usually estimated by calculation with mathematical formulae. Three-dimensional body scanning (3D scan) offers a suitable alternative.
OBJECTIVES:
We determined the BSA in healthy term and near-term neonates by 3D scanning. This system should be useful in the setting of intensive care medicine.
METHODS:
The measuring system consisted of a projector, two cameras, mirrors and a computer, and used the fringe projection technique with visible light. The infants were examined in a supine position; the hidden parts of the bodies were corrected for using a mathematical factor developed with a baby doll model. Results of the 3D scans were compared with those from five mathematical formulae for each subject.
RESULTS:
A total of 209 infants were studied by 3D scanning, of whom 53 had acceptable images and were selected for further analysis. The mean BSA was 2,139 cm(2) (SD 223.72). The minimal BSA was 1,587 cm(2), the maximal 2,670 cm(2), with a good correlation to body weight and length. One mathematical formula (Du Bois and Du Bois) showed a distinct underestimation of BSA compared to 3D scanning, the others an overestimation. Mean percentage similarity was from 96.8 to 100.9%.
CONCLUSIONS:
3D scanning is an accurate and practical method to estimate BSA in newborns. Individual and repeated measurements from day to day are possible. Further studies are warranted in preterm and sick neonates.
Copyright © 2011 S. Karger AG, Basel
Excellent find! Thank you so much!
From what we're seeing with our own estimates, their results don't surprise me too much, but it's great to have some confirmation that what we think we're seeing is replicated elsewhere!
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Physical Phenomena
- Luminescence
- Fluorescence