Interpretation of Diagnostic Images

What is the 3-Stage Approach to analyzing a radiograph?

Stage 1 - Recognition

• Perform a systematic search of the image (identify what’s normal)

Stage 2 - Report

• Identify and describe any abnormalities

Stage 3 - Interpretation/Analysis

• Make a diagnosis, then generate a systematic differential diagnoses list

  • Most likely first, least likely last

• Plan for further investigation/treatment

Describe Stage 1… what is one key point of this approach?

Systematically search the ENTIRE image. Can go either:

• System by system (cardio, musculoskeletal, digestive, etc.)

• Area by area (outside-in, left-right, etc.)

Key point: READ THE WHOLE IMAGE

• Don’t get distracted by obvious lesions!!

• Also, remember to take 2 views always; helps us see depth (see image)

Describe Stage 2… What are some changes you may see in a radiograph?

Stage 2 is for identifying abnormalities. To do that:

• NEED TO KNOW WHAT NORMAL LOOKS LIKE

Describe any changes you may see. Changes in:

• Size

  • Abnormal increase (hypertrophy, neoplasia, etc.) or abnormal decrease (atrophy, hypoplasia)

• Shape / margins

  • Either altered shape/margin (via trauma, necrosis, neoplasia, etc.) or loss of visibility overall (ascites)

• Number

  • Either altered number of normal structure (increase or decrease) or the number of lesions (primary vs. secondary disease)

• Position / location

  • Abnormal position or displacement (due to an adjacent mass, organ enlargement, torsion, hernia, or ectopia)

• Opacity

  • Either increased (fluid/soft tissue accumulation, mineralization) or decreased (air accumulation, decreased bone density, fat accumulation)

• Internal structure (advanced imaging)

  • Ex. MRI of brain with brain tumor

• Function (seen via contrast xrays)

  • Ex. Kidney function using contrast

Describe Stage 3… how would you put together a list of differentials based on your findings?

Generate differentials for ALL abnormalities found

• Some may be incidental

Create a structured list; most likely first

• Differentials that explain ALL changes should be ranked higher

• Consider other clinical findings from the patient’s history; may influence differential list

What is the difference between an “informed interpretation” of an image, and a “blind” interpretation of an image?

Informed Interpretation:

When you are aware of the patient's history (and other findings) prior to image interpretation. Allows for:

• Better ability to recognise and dismiss incidental lesions

• Earlier refinement of differential lists

• Risk of bias

  • Expecting specific findings

  • Over-reading images

  • Temptation to finish examination when suspected lesion found

• Most commonly used in practice

Blind Interpretation:

When you have no information on the patient (history) prior to image interpretation. Allows for:

• No bias to particular disease

• Less likely to terminate examination early

• Longer and less structured differential lists

  • Could potentially cloud clinical picture

• Most commonly a second opinion

BEST TO DO A MIX OF THESE; perform your own informed interpretation, but then have a colleague do a blind interpretation and see if they agree with you

What are the 3 most common errors in image interpretation?

1) Search errors

2) Judgement errors (under or over-reading)

3) Analysis errors

How are search errors in film interpretation caused?

BY NOT LOOKING AT EVERYTHING

• Missing lesions

  • Either not seen, interpreted as normal, or failing to examine the entire image

    • Can then lead to early termination of the search after finding a “significant” lesion or obvious change

    • Ex: patient comes in with shoulder pain; xray of joint itself looks fine, so no further investigation is done on the image. Meanwhile, there was a osteosarcoma on the scapular spine that was missed because the vet didn’t look at the image as a whole

• Not noting absence of a normal structure

• Inadequate image quality

How are under-reading errors in film interpretation caused?

• Failure to identify lesion

  • Poor quality image

  • Inexperience

• Failure to consider appropriate differentials

  • Preconceived ideas

    • Expecting a particular diagnosis before looking at the image

    • Get a “blind” film reading?

  • Identifying a certain combination of changes as a specific diagnosis

  • Failing to reconsider differentials when presented with new evidence

How are over-reading errors in film interpretation caused?

IDENTIFYING A NORMAL FEATURE AS PATHOLOGY

• E.g. end-on pulmonary vessel read as nodule

• E.g. retraction of pulmonary edges in fat animals read as pleural effusion

• E.g. nutrient canal in bone read as fracture

NOT RECOGNIZING BREED VARIATIONS

FAILURE TO RECOGNIZE NON-DIAGNOSTIC IMAGES

• Mistaking image artifact for pathology

MISTAKING INCIDENTAL FINDING FOR PATHOLOGY

How are analysis errors in film interpretation caused?

Generation of an INCORRECT or INCOMPLETE DIFFERENTIAL LIST

• May be linked to a fault in the search and report stages of your interpretation

• May be due to preconceived ideas about a likely diagnosis

Inappropriate further investigation

• Failure to recommend necessary investigations

• Recommendation of unnecessary investigations