Do I need to calibrate images?

Support
#1

Do I have to calibrate images to be able to measure? I will sometimes get radiographs (DICOM no jpeg) from referring veterinarians that do not have a calibration device in them. Is there a way that I can still make rough measurements on them (understanding that calibrating the images would make the measurements more accurate)?

Thanks,

Sebastian

#2

Yes, you do need to calibrate against a marker that is at the same height as the area of operation to avoid any perspective skew.

DICOM modality calibration can be inaccurate so we suggest always calibrate manually. There have been discussions to make this compulsory in vPOP even if there is pixel spacing metadata in the DICOM.

If there are no markers of a known size to measure against, then there is not much possibility of measuring to any meaningful accuracy.

Kind Regards

Nick

#3

Hi Nick, I understand that it would be ideal to calibrate the image. Unfortunately that is not always possible for me. I am a mobile surgeon and we go to over 100 different clinics. It would be a lengthy process to train all the technicians at all those clinics to include a calibration device in their radiographs.

If a lateral stifle radiograph is taken with the limb against the table, I understand that there will be some distortion of the image, but it has not been big enough to cause a problem for me in planning my surgeries.

I really like VPOP, but if I can only make measurements on images that can be calibrated, it is going to be very challenging for me to use. )-:

On a different subject, is there an app for VPOP for Androids?

Thanks again,

Sebastian

#4

Good day Sebastian, a pleasure to e-meet you. If I can chime in here as a mobile surgeon, I can entirely relate to the challenges you face and can make the following suggestions.

  1. vPOP ’ just needs’ a reference. So if you need to measure an angle such as a TPA, you can enter a value between two landmarks of your choice and tell vPOP what you believe them to be. eg , if you are cutting ‘00s of tplos a year, you will know through experience what your patient signalment will be. eg intercondylar eminences to PatTend insertion … within that , a 15mm blade will fit a Westie, a 21mm for a lab, 24mm large lab. You can then set your target 5’ TPA and rotate your prox frag to it, or use the charts . The only thing you cannot do of course, is to measure your absolutes- D1, 2 etc. On the day of surgery, you could either measure the PT insertion to the lateral malleolus and recalibrate that by applying it to your plan, or take a preop rad to your spec and positioning with your marker.

2)Rather than training up all your techs in your host clinics- I can highly recommend having a metal business card with left/ right and linear measurements to give to your clinics. It won’t degrade, it’s highly useful to your hosts and they know who to call when they get a case!

  1. Train your hosts in radiographic positioning/ best practice in your absence by giving them a read only vPOP case share of photographs of how you like your patients positioned WITH A MARKER. eg my own personal one for thoracic limb positioning/ calibration. I simply share this weblink by any messenger or email for an instant resource to rDVMs and Techs. vPOP.

Bottom line- you cannot rely on Dicom Metadata due to the magnification error/ variation created by the differing distances of the anatomic area to the plate be it within a patient or across the potent breed sizes of our patients.

I hope the above is useful. kind regards RP

1 Like
#5

For the question regarding Android. You can use the browser version at https://app.vpop-pro.com

I hope Rory’s answer covers your other concerns.

Kind Regards

Nick