About
What is Reference Line – Training and Evaluation?
Reference Line – Training and Evaluation is a webapp designed to train and evaluate HR-pQCT operators when positioning the reference line on scout view images.
Reference Line – Training and Evaluation contains three modules: “Training”, “Evaluation”, and “Repro”. Each module simulates the acquisition software of the XtremeCT system (Scanco Medical AG, Brüttisellen, Switzerland), reproducing visualization of scout view images and handling of reference line positioning.
During a scan, HR-pQCT operators acquire an anterior-posterior 2D projection (scout view image) of radius and tibia, where they visually identify an anatomical landmark that determines the region to be scanned. The landmarks correspond to a peak in the mid-joint for radius, and to a peak in the distal articular plateau for tibia, as indicated by the black arrows in the figure below. Operators manually position a reference line at the landmarks (green line), determining the region to be scanned (green rectangle), which is located distally in the standard protocol.
Visual identification of anatomical landmark and positioning of reference line can vary for a given operator as well as between operators, affecting location of the region of interest, and thus measurements of morphological and densitometric parameters. To overcome these inconsistencies, we propose a tool to train operators at identifying anatomical landmarks and positioning the reference line, in order to minimize intra- and inter-operator variability.
If you are using Reference line – Training and Evaluation to train operators that take part in your study, please cite us in your paper as:
Bonaretti S., Vilayphiou N., Chan C.M., Yu A., Nishiyama K., Liu D., Boutroy S., Ghasem-Zadeh A., Boyd S.K., Chapurlat R., McKay H., Shane E., Bouxsein M.L., Black D.M., Majumdar S., Orwoll E.S., Lang T.F., Khosla S., Burghardt A.J. Variability In HR-pQCT Scan Positioning Is Significant in Whole Acquisition Precision and Decreases After Training. Submitted to Osteoporosis International. 2015.
Documentation
To identify the anatomic landmark for positioning the reference line in distal radius exams, we focus specifically on the anatomy of the articular surface of the radius in the radiocarpal joints of the wrist.
The radius landmark is a peak of the radiocarpal articular surface of the radius. It separates the radiopaque subchondral bone (bright signal) from the radiolucent joint space (dark signal) (see figure below).
In simple cases (a), where a peak is clearly visible (blue dot), we position the reference line (green line) so that it intersects the peak of the crest at the edge of the joint surface. In more complicated cases, where there is no visible peak in the scout view images (b), we consider the mid-point of the curve that traces the articular surface of the radius (blue dot at the midpoint of the red line). For both cases the articular surface is identified as the edge in the image where there is a clear transition between the radiolucent (dark signal) joint space and radiopaque (bright signal) subchondral bone. The blue lines tracing the articular surface directly overlay that edge. The reference line is positioned so that it intersects this curve.
When a crest is present in the scout image, we can use the Swiss hiker analogy to describe the correct location to intersect this curve (see figure below). Imagine a hiker walking along the surface represented by this curve. The crest is like a small hill that he must climb over as he treks across the joint surface. At the peak of this hill he wants to plant a flag to commemorate his accomplishment. The location where the hiker plants the flag is analogous to the point on the curve that must be crossed by the reference line, when correctly positioned.
To identify the anatomic landmark for positioning the reference line in distal tibia exams, we focus specifically on the anatomy of the articular surface of the tibia at the tibial platfond.
The tibia landmark is a peak or the crest of the tibia platfond, which separates the radiopaque subchondral bone (bright signal) from the radiolucent joint space (dark signal) (see figure below).
When the tibial plafond is flat (see the blue line in the figure below (a)), we position the reference line (green line) so that it overlays the thin, high-intensity plateau of the plafond. When there is a crest in the tibial plafond (b), we identify the landmark as the peak of the crest in the tibial plafond (blue dot), and we position the reference line (green line) so that it intersects the thin, high intensity margin at this peak of the tibial plafond. Note that for the tibia the reference line cuts through the center of the high intensity plateau rather than the surface crest as in the radius.
The training tool was specifically designed to allow HR-pQCT operators to train for positioning reference line on scout view images.
You can train using scout view image sets which each have 8 images. Each set has a category, which you can select from the following options - limb laterality (left/right) together with landmark visibility (visible/no visible), mixed sets (combined literalities and landmark visibility) and random sets (any of the above possibilities). There are 3 sets of each category, which are randomly presented to the user. Before moving to the evaluation module, complete the set "For certification," which contains all available sets.
The evaluation tool was specifically designed to evaluate HR-pQCT operator performances and to certify operators.
After training on the "For certification" sets of radius and tibia images, the HR-pQCT operator can move to the evaluation step. To be certified, you must pass at least 3 sets for the radius and 3 sets for the tibia. In order to pass, the sum of “Exact” and “Good” positioning has to be 7, thus the number of “Out of range” positioning is at most 1. If you have difficulty with succeeding, you can return to the training tool for more practice before starting a new evaluation session.
The reproducibility tool was specficially designed to evaluate HR-pQCT intra-operator reproducibility.
The reproducibility tool consists of 15 images of the radius and 16 images of the tibia. For each bone, images are repeated randomly twice for a total of 45 images for the radius and 48 for the tibia. After completing the entire set (of either radius or tibia images), you will see a table that summarizes your positioning for all the images and indicates if your reproducibility is in the same range as that of experienced operators.