X-RAY SCORING SYSTEM

This tutorial aims to introduce you to a scoring system to evaluate x-ray images of keel bones in laying hens. Further it provides a training session which intends to standardize the scoring of keel bone x-rays.

Part I: INTRODUCTION

Welcome!

The Scoring System

Thanks for your interest in our x-ray scoring system. This tutorial includes

  • background information
  • introduction to the scoring system
  • training session

For this tutorial, please print out the two following files (format A4, no scaling):


One more thing!

Even if you're used to scoring x-rays and / or keel bones, please try to get yourself into this tutorial and training session and follow the instructions carefully.

First: What is a fracture?

Fractures on x-ray images

Before introducing you to the scoring system, I would like to give you a short introduction about x-ray imaging and how fractures are defined from a radiologist's perspective.

Depending on the density of the structures (bones and soft tissue) being x-rayed, the structures appear in different brightness levels on the image. The higher the density of the structure, the brighter it appears on the image.

From a radiologist's perspective a fracture is defined as a visible fracture line / fracture gap. As this means a separation of the bone, a fracture appears as a dark line on the image (no x-rays being absorbed by the bone at the fracture gap). During the healing process of the fracture, bone material will accumulate around the fracture line and form a callus. As the callus is an accumulation of bone, it has a higher densitiy than healthy bone and will therefore appear brighter on the image. Additionally, the shape of a callus is usually visible on x-rays as a deviation of the straight keel contour, especially if dislocation of the fracture site was involved.

The aim of this tutorial is not to determine healing states of fractures. However, it is important to know how fractures are defined - especially because the definition of fractures differs greatly between x-rays and palpation. On x-rays, even small and very new fractures can be detected whereas palpation focuses on the detection of calluses and therefore healing or healed fractures.

On the images below, I would like to show you the difference between fractures (= fracture gap visible) and healed fractures. Thereby, it's important to remember some methodological constraints:

  • for a valid and detailed description of fractures and structural changes of the bone, at least two images from different perspectives would be needed (instead of latero-lateral only)
  • a detailed description of the lesions would require a special analytics software and appropriate equipment (e.g. high resolution screen)
  • higher x-ray density could also result from structural changes (e.g. sclerosis) rather than from a fracture. In order to determine a callus as a result from a fracture, longitudinal x-rays are needed to identify initial fractures.

In the images below, blue circles indicate visible fracture gaps and green circles mark calluses / structural changes.

Example 1

One visible fracture gap in the lower half of the bone; around the dark line, white shadows indicate the accumulation of bone material (= initation of healing process) but no callus is visible yet.

Two calluses in the lower part of the keel; no fracture gap visible, but clear deviation from the straight line on the dorsal site of the keel (right).
Especially the upper callus is very likely to be the result of a fracture because a slight dislocation / angle of the lower part can be identified. 

Example 2

Two visible fracture lines on the upper part of the keel, both with white shadows on both sides indicating the start of the healing process.

One callus visible at the tip on the dorsal side of the keel. As there is no dislocation or angle visible, longitudinal images would be needed in order to state if this structural change has been a fracture initially.

Example 3

Two fractures gaps visible at the tip of the keel, no healing process started yet.

One callus with clear dislocation / angle of the lower part of the keel, but no fracture gap visible anymore.

Example 4

Two fracture gaps visible at the top of the keel, whereas healing hasn't started yet (no white shadows on the edge of the bone fragments).

Several calluses at the tip of the keel. There seems to be dislocation and a clearly visible edge on the ventral side (left). The smaller, multiple calluses on the dorsal side (right) could be the result of compression rather than fractures. Again, longitudinal images would be needed to make a statement about initial fractures.

Aim of the scoring system

Aim & reliability

Both inter- and intra-observer reliability of this scoring system were found to be high (ICC inter = 0.985, ICC intra = 0.923) after completion of the training session, irrespective of background and experience of observers. Therefore, the scoring system is supposed to be applicable for all interested people working with x-rays of keel bones. Although there are not too many groups performing x-rays of keels, a standardized evaluation leads to comparable results.

Studies using x-rays will serve to answer different research questions, but all of them will need to quantify the severity of fractures. I therefore propose to use two variables:

  • One variable describing the severity of fractures = topic of this tutorial
  • One variable describing the mechanism of interest (e.g. pain, healing process) = depending on research question, therefore NOT topic of this tutorial.

The two variables I

The two variables

Variable I: severity

Using x-rays instead of palpation enabled us to describe the variety and complexity of keel bone fractures in great detail rather than using a simplified palpation score. X-rays allow us to assess:

  • Number of fractures
  • Location of fracture(s)
  • Fracture type(s)
  • Fracture direction(s)
  • Width of fracture gap(s)
  • Angle(s) / dislocation(s)
  • Callus formation
  • Duration of healing
  • ...

All the measures above are easy to take on single lesions, but are difficult to summarize / add up to a single value for each keel bone of hens suffering from multiple fractures. In our first set of x-ray images, we found hens with up to 15 fractures / lesions, all of them being different regarding type, healing state and other characteristics. Further, grouping for specific characteristics (e.g. fracture type; comminuted fractures vs. other fractures types) is difficult because too many different combinations of type, number, location and other characteristics of a fracture exist. Regarding our experimental design, it is crucial to determine one single score per bone instead of scoring fractures and lesions individually.

Existing studies didn’t need to consider the complexity of keel bone fracutures to this extent because palpation couldn’t provide as much information as x-ray images. I therefore propose a scoring system to score severity of fractures which allows to evaluate a keel bone in its entirety. Additionally, the system offers the possibility to not only include severity of fractures, but also account for mechanisms of interest.

The two variables II

The two variables

Variable II: mechanism of interest

We all agree that keel bone fractures (and fractures in general) are likely to affect an individual in different ways, initiating different mechanisms. Depending on the research question, a specific mechanism could be relevant regarding the effect of fractures: for mobility, pain and resulting behavioural changes might be the main mechanism whereas pain might be rather irrelevant when investigating productivity. Instead, the healing process and resulting metabolic changes (e.g. increased calcium / energy demand) are more likely to affect productivity. It is therefore obvious that we need to consider the ongoing mechanism which is relevant for our research question besides taking into account the severity of fractures.

Examples:

  • Productivity: callus formation / bone healing (e.g. «active bone» vs. «inactive bone»)
  • Pain: moveable bone fragments (e.g. «fracture gap visible» vs. «no fracture gap visible»)

This variable could be assessed completely separate from the severity of fractures and is NOT the topic in this training session. However, it’s important to know that not only severity, but also the mechanism of interest could and should be considered in the proposed scoring system. Using two variables will both quantify (variable I: severity) and describe (variable II: mechanism of interest) the ongoing process.

To summarize: variable II depends on the research question, aims to characterize ongoing mechanisms and is NOT topic of this tutorial.

How to score severity?

Back to severity...

As mentioned before, the aim of this severity scoring system is to determine one single severity score per bone and thereby accounting for the whole variety and complexity of multiple fractures. I therefore propose to use a tagged visual analogue scale.

The tagged visual analogue scale will not only provide a tool to assess the complexity of multiple fractures and therefore enable to score a bone in it's entirety, but will further be a continuous measure quantifying the mechanism of interest when using an interaction term in the statistical analysis (severity * mechanism; see figure below).

Figure: Interaction of severity and mechanism of interest

Overall, we would expect a new fracture to be more painful than a healed fracture. Increasing severity will probably increase pain in both cases, but increase in pain is likely to be greater in new fractures than healed fractures.

Tagged visual analogue scale

Tagged visual analogue scale

The proposed scale ranges from 0 (= no fracture) to 5 (= extremely severe). It is a continuous measure (everywhere between 0 and 5) but provides additional 4 tags to help you to orientate. Please keep in mind: these tags are artificial, it was me who tried to find 4 intermediate scores between 0 and 5 and selected only one image per score which I thought would represent it the best.

If printed on A4, the scale is exactly 10 cm long. For scoring, you can just tag the scale wherever you think it's appropriate. The exact value can then be extracted by just measuring the distance between the left end (0) and your tag.

"example scores"

"example scores"

The scale is not the only tool I will provide for scoring. As mentioned before, the images in the final scale are chosen by me – but fractures can be so complex and diverse that these few images can’t describe the scores detailed enough. On the other hand, words and numbers can't describe the complexity of fractures either, which is why I would like you to use the "example scores".

These example scores you can find in the next slides are meant to serve as a «catalogue». I basically used these example images instead of words and numbers to describe the scores and the scale. The example scores should give an overview of e.g. «what could be a 3» – but remember: the scale is continuous, so the example scores are only proxies too (e.g. score 3 could range from 2.5 – 3.49). Feel free to compare them with the images I’ve selected for the final scale.

In the following slides, I will show you each score with the example images I've assigned. I won't describe the scores in words and numbers on purpose because there will always be images which won't match these descriptions. When looking at the images of the example scores, please try to focus on the following:

The score is

  • not only based on the number of fractures 
  • not purely location dependent 
  • not purely type dependent (dislocation / angle, fracture gap, ...) 
  • supposed to estimate «amount of bone affected» and to score the severity of multiple fractures in their entirety.

Score 0

Score 0

Please study the images above and focus on

  • Number of fractures
  • Location of fracture(s)
  • Type of fracture(s)
  • "amount of bone affected"

Score 1

Score 1

Please study the images above and focus on

  • Number of fractures
  • Location of fracture(s)
  • Type of fracture(s)
  • "amount of bone affected"

Score 2

Score 2

Please study the images above and focus on

  • Number of fractures
  • Location of fracture(s)
  • Type of fracture(s)
  • "amount of bone affected"

Score 3

Score 3

Please study the images above and focus on

  • Number of fractures
  • Location of fracture(s)
  • Type of fracture(s)
  • "amount of bone affected"

Score 4

Score 4

Please study the images above and focus on

  • Number of fractures
  • Location of fracture(s)
  • Type of fracture(s)
  • "amount of bone affected"

Score 5

Score 5

Please study the images above and focus on

  • Number of fractures
  • Location of fracture(s)
  • Type of fracture(s)
  • "amount of bone affected"

Now: training!

Training instructions

Tutorial is over! Now you have looked at the example scores in more detail. The next section of this course aims to train you to become more familiar with each of the 6 scores (0 - 5). I will show you 64 images from the example score catalogue which you should score 0 - 5 (single choice). You will get immediate feedback - if you scored correctly, the next image will appear. If you were wrong, you can try again. If you needed more than one attempt, please check the example scores catalogue in order to figure out why you scored incorrectly first.

For training, you won't use the continuous scale yet. Training is supposed to give you a feeling for the ranges of each score (e.g. score 3 ranging from 2.5 - 3.49) and to show you again how different a specific score can look like. As you will score only images from the catalogue, you won't be allowed to use the example scores for the training session unless you needed more than one attempt to select the correct answer. Instead, you can use the scale with one selected image for each score.

Please click on "home" and select the next section "training session" to start with the training.

Part II: TRAINING SESSION

Training instructions

Some reminders...

Before you start with the training session, please remember:

  • The score is supposed to quantify the «amount of bone affected» and to score the severity of multiple fractures in their entirety
  • Make sure to «scan» the whole bone – there might be fractures not only at the tip, but also on the top of the keel!
  • Please only use the scale without having a look at the catalogue of example scores
    (for scoring, you will be allowed to use both the tagged scale as well as the catalogue of example scores)
  • If you needed more than one attempt, please check the example scores catalogue to figure out why you scored incorrect initially.

Please click "next" to start with the training session. There are 64 images to score.

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

please score!

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5

Training completed!

Training completed - and now?

You're done with the inital training session. In case you would like to re-train, you can either start the tutorial again (new browser session, no login) or use the recalibration tool. We recommend to recalibrate with this tool periodically (e.g. once a month, after long breaks) to ensure that the initial protocol is maintained. 

A small reminder:

You only trained on "full scores", but the scale offers the opportunity to get a continuous measure of severity by tagging the 10 cm line. You could use the following sheet with empty scales (print out in 1:1 ratio to make sure that the scale is still 10 cm):

Tipp

However, printing on paper and measuring each line is time consuming. I'd therefore recommend to create something like this:


I’ve just printed the scale and put it into a plastic file, then glued a 10 cm ruler on the file and used a piece of plastic with a tag on it to mark the scale. This enables you to use the scale with the images, read the value from the rulerand enter it in a table directly.

Thank you!

I hope this scoring system tutorial will be helpful for your research.

Feel free to contact me with more inputs or questions: christina[email protected]