Clinical history.
The dog was first seen for chronic lameness of the right hind which had become progressively worse over the last two weeks. Manipulation of the limbs revealed right hip pain. The dog was booked in for radiography of the pelvis.

Restraint for radiography.
This was achieved by general anaesthesia. 

Choice of film/screen/grid.
Agfa CP-G Plus (30 cm x 40 cm) with a Curix rare earth screen. This is a standard green sensitive general purpose film.

A stationary parallel grid was used as the tissue depth was greater than 10 cms.

Positioning.
Ventro-dorsal pelvis.
Positioning for this projection was achieved by placing the dog in dorsal recumbency with the thorax supported in a cradle. Both stifles were inwardly rotated and secured with sellotape, and the patellae were palpated to ensure central location between the femoral condyles. Tapes were attached to both feet and the hind legs extended and tied to the table to ensure both femora were parallel to the table top. The dog was checked to ensure it was placed in a bilaterally symmetrical position. See picture right, showing positioning for the dorso-ventral projection of the pelvis. A metal right marker was placed on the grid.

Safety procedures undertaken during radiography are described in Appendix 1.

Centering and collimation.
The primary beam was centered on the cranial border of the pubic symphysis, and collimated to include the patellae and all the pelvis.

The following exposures were used:

Radiographic appraisal.

Positioning
The pelvis is slightly rotated to the left - the left wing of the ilium appears larger than the right and the sacro-iliac joint space is larger on the right. Usually the obturator foramen is smaller to the side of rotation but it is difficult to see on this radiograph. The spine is not quite straight and curves to the right. This dog proved difficult to position as the right hip would not extend fully, despite general anaesthesia - see Figure 1. This might account for the rotation. The patellae are both centrally located on the femora.

Centering
The primary beam was centered on the pubic symphysis approximately 3 cms caudal to the cranial border of the symphysis, whereas it should have been on the cranial border. This is a relatively small discrepancy but palpation of the femoral heads may have improved the accuracy of centering, as the femoral heads lie in line with the cranial border of the pubic symphysis.

Collimation
All four edges of the radiograph are unexposed, demonstrating the extent of the primary beam. Collimation could have been closer laterally. This could be improved by palpating the bones instead of relying on soft tissue for collimation of the area. It would be difficult to collimate any closer cranially and still retain centering and caudal collimation in the correct position.

Exposure
The exposure of this radiograph is satisfactory. The contrast is good, having a large range of well differentiated densities within the bones and the soft tissues can also be seen. Detail is also of a reasonably high quality on this radiograph. Fine detail relies both on good contrast and good definition. Several factors are involved in obtaining sharp definition. Restraint by general anaesthesia improves definition, as movement is reduced. Using a grid when tissue depth is over 10 cms (15 cms for the thorax), when higher kVs are required to penetrate tissue, will also help improve definition. This is due to the grid slats absorbing secondary radiation and only allowing X-ray photons which are travelling in the same direction as the primary beam to pass through the grid, resulting in a sharper image. Use of a lead mat or lead lined table to absorb X-ray photons will reduce back scatter and improve definition. Correct centering and close collimation will also have an effect on the sharpness of the image by reducing scatter and ensuring the area of interest has the smallest distortion possible - the centre of the primary beam has the least divergence of X-ray photons.

Artefacts
There are several on this film:

• A grid artefact appears as a mottled area above the right ilium.
• There are parallel scratch lines running the length of the film, caused by the rollers in the automatic processor. This could be resolved by either more frequent cleaning of the rollers or replacing them if they are damaged.
• The grid lines are visible on the radiograph - this is usual when using a stationary grid.
• There are several finger prints on the radiograph, caused after processing.

Processing
This radiograph was processed automatically, and is correctly developed - the background would be grey if under developed, unless grossly under exposed. If the radiograph had been over developed the metal marker would be grey instead of white. Note - there are grey areas on the marker but this is due to the marker not lying flat on the grid and, as a result, being partly exposed.

General comments
This is a diagnostic radiograph for the purpose it was intended. A diagnosis was made of osteoarthritis of the right hip, possibly secondary to hip dysplasia although it was noted that the left hip did not show any gross abnormalities which is unusual in cases of hip dysplasia, as the condition is usually bilateral. German Shepherds are the breed most susceptible to hip dysplasia.

The treatments available for this condition would be conservative or surgical - by performing an excision arthroplasty, total hip replacement or pelvic osteotomy. In this case the treatment chosen by the veterinary surgeon was conservative with phenylbutazone 200 mg tablets, one to be given twice daily with food. This decision was based on the weight being considered too heavy for successful excision arthroplasty, the cost of a hip prostheses being beyond the client's budget and the joint having too many degenerative changes for a pelvic osteotomy to be performed.

The purpose of this radiograph was to obtain a medical diagnosis, but this would be the required projection, positioning and collimation (including the patellae) necessary for a pelvic radiograph which was to be submitted for hip dysplasia certification. However, it would be necessary to record the date, left/right marker and dog's kennel club number using X-RITE tape, or similar, to appear permanently on the radiograph at the time of processing. Very accurate positioning is also necessary for hip scoring.

General assessment of radiography
Radiography was useful in this case for the following reasons:

• It is a non invasive procedure, and diagnostic radiographs could probably have been achieved with sedation, although better positioning and definition can be obtained under general anaesthesia.
• A radiograph enables sight of bone changes, which are not always apparent to the naked eye.

References:
Principles of Veterinary Radiography, Douglas, Herrtage & Williamson, Bailliere Tindall, 4th Ed. 1987 Ch 4, 7, 9.
Manual of Small Animal Diagnostic Imaging, BSAVA 2nd Ed 1995 Pg 124, 125, 183