QUANTIFICATION OF THE EFFECT OF VARIOUS PATIENT AND IMAGE FACTORS ON ULTRASONOGRAPHIC DETECTION OF SELECT CANINE ABDOMINAL ORGANS
Presented at the IVRA meeting, Vancouver, Canada, August 2006.
Dr. Nathalie Rademacher's present address is Department of Radiology, School of Veterinary Medicine, Louisina State University, Baton Rouge, LA 70803.
Abstract
We assessed factors that affected ultrasonographic visualization of the pylorus, duodenal papilla, pancreas, adrenal glands, and jejunal and medial iliac lymph nodes in the dog. An abdominal ultrasonographic examination was performed on 100 canine patients, equally divided between two facilities. The pylorus was visible in 64% of the dogs, the major duodenal papilla in 42%, the left pancreatic lobe in 56%, the body of the pancreas in 60%, the right pancreatic lobe in 87%, the left adrenal gland in 91%, the right adrenal gland in 86%, the medial iliac lymph nodes in 54%, and the jejunal lymph nodes in 51%. The parameters that negatively influenced the visibility of these organs were the presence of air or food in the gastrointestinal tract (pancreas, duodenal papilla), age (lymph nodes), and body weight (pancreas, duodenal papilla). The parameters that positively influenced their visibility were the presence of air or food in the gastrointestinal tract (lymph nodes), body weight (lymph nodes), body condition score (right adrenal gland), and inherent image quality (left pancreatic lobe). There was a significant difference between the two institutes for the visualization of the pylorus, pancreas, and lymph nodes, which was probably related to different body positions used for scanning in each institution.
Introduction
Multiple parameters such as obesity, gastrointestinal contents, incompliance, and thoracic conformation can hamper sonographic visualization of some abdominal organs in the dog, such as the pancreas, adrenal glands, lymph nodes, and pylorus.1–10 However, the absolute detection rate of these abdominal organs, and how this is affected by various parameters, is unknown. Thus, the aim of this prospective study was to quantify the influence of age, gender, body weight, body condition score, patient compliance, image quality, and gastrointestinal content on organ visibility.
Materials and Methods
An abdominal ultrasonographic examination was performed in 100 client-owned dogs between November 2005 and January 2006. Fifty dogs were examined at Ghent University and 50 at the University of Zurich. All the abdominal sonograms were performed by ECVDI residents in training or board-certified radiologists.
At Ghent, examinations were performed with a microconvex (6–10 MHz) and a matrix linear (7–14 MHz) transducer.* The dogs were mainly in dorsal recumbency. Lateral recumbency was used only when the animal was not cooperating or for evaluating specific organs. At Zurich, dogs were imaged using a curvilinear (5–8 MHz) and a linear (5–12 MHz) transducer.† Dogs were almost always in right recumbency, with dorsal recumbency rarely being used.
The dogs were patients undergoing an ultrasonographic examination for a variety of complaints, not always related to the abdomen. Dogs that had obvious sonographic abnormalities of the organs selected were not included in the study. Age, breed, gender, and body weight were recorded. A body condition score from 1 to 5 was assigned (1=very thin, 5=obese). The image quality inherent to the animal was graded from 1 to 5 (1=extremely poor, 2=poor, 3=moderate, 4=good, 5=excellent). It was noted whether the dog was fasted for at least 12 h. The gastrointestinal content (air or food) was recorded and graded from 1 to 5 (1=extremely poor image quality because of a large amount of gastrointestinal gas or food, 5=excellent image quality because of a very small amount of gas or food). Finally, the compliance of the animal was recorded without and, if needed, with sedation. Without sedation, a grade from 0 to 5 was assigned (0=sedation required, 1=very difficult, 2=difficult, 3=moderate, 4=good, 5=compliant). The same grading scale (1–5) was used to assess the compliance with sedation.
During the ultrasonographic examination, the observers were asked to assess the following organs: pylorus, major duodenal papilla, pancreas (body, left and right lobes), adrenal glands (left and right), and lymph nodes (jejunal and medial iliac). For each organ, the observers recorded whether it was visible and, if yes, whether its size was within normal limits.
The percentage of visualization of each organ was determined. A logistic regression model was used to determine the influence of patient and image parameters and the institutes on organ visibility. All analyses were performed at the 5% significance level.
Results
The mean age was 7.8 years (range: 0.5–14 years). Nine dogs were 1-year-old or less. There were 37 intact males, 14 neutered males, 17 intact females, and 32 neutered females. The mean weight of the dogs was 21.5 kg (range: 1.4–51 kg). There were 12 dogs belonging to dwarf breeds (<5 kg), 17 to small breeds (5–10 kg), 19 to medium breeds (10–20 kg), 42 to large breeds (20–40 kg), and 10 to giant breeds (>40 kg). Six dogs were very thin, 10 were underweight, 43 had an ideal body weight, 34 were overweight, and 7 were obese.
The inherent image quality and gastrointestinal filling data are summarized in Table 1. Twenty-nine dogs were not fasted and the remaining 71 dogs did not receive any food within the previous 12 h. The compliance of the dogs without and with sedation is summarized in Table 2.
| Image Quality | Inherent (%) | GastrointestinalFilling (%) |
|---|---|---|
| Extremely poor | 0 | 3 |
| Poor | 15 | 22 |
| Moderate | 25 | 32 |
| Good | 41 | 34 |
| Excellent | 19 | 9 |
| ComplianceWithout Sedation(n=87) | Compliancewith Sedation(n=13) | |
|---|---|---|
| Very difficult | 4 | 0 |
| Difficult | 11 | 0 |
| Moderate | 19 | 1 |
| Good | 30 | 5 |
| Compliant | 23 | 7 |
The frequency of visibility of each organ is summarized in Table 3. The influence of variables related to the dog (age, gender, body weight, body condition score, inherent image quality, gastrointestinal content, and compliance with or without sedation) on the visibility of organs was quantified using the odds ratio (OR). The compliance of the dogs and gender had no significant influence on the visibility of the abdominal organs of interest.
| Organs | Total Visibility |
|---|---|
| Pylorus | 64/100 |
| Duodenal papilla | 42/100 |
| Pancreas: body | 60/100 |
| Pancreas: left lobe | 56/100 |
| Pancreas: right lobe | 87/100 |
| Left adrenal gland | 91/100 |
| Right adrenal gland | 86/100 |
| Jejunal lymph nodes | 51/100 |
| Medial iliac lymph nodes | 54/100 |
There was a significant difference between the two institutes with regard to the visualization of the pylorus, with 94% visibility in Ghent and 33% visibility in Zurich (P<0.0001). No other parameter influenced visualization of this organ.
Increasing body weight and presence of gastrointestinal contents negatively influenced visualization of the major duodenal papilla (OR=0.95, P=0.007 and OR=2.00, P=0.002, respectively).
There was a significant difference between the two institutes with regard to visualization of the pancreas (Ghent 59% and Zurich 76%, P=0.048). Gastrointestinal content had a negative influence on the visibility of the pancreatic body (OR=2.63, P<0.0001), the right pancreatic lobe (OR=4.35, P<0.0001), and the left pancreatic lobe (OR=1.67, P=0.015). The visibility of the pancreas was negatively influenced by body weight (body: OR=0.96, P=0.023; right lobe: OR=0.93, P=0.005; left lobe: OR=0.95, P=0.004). The visibility of the left pancreatic lobe was positively influenced by the inherent image quality (OR=2.00, P=0.002).
No parameter influenced visibility of the left adrenal gland. The visibility of the right adrenal gland was enhanced in dogs with a high body condition score (OR=2.17, P=0.011).
Lymph nodes were identified more often in Ghent than in Zurich (Ghent 76% and Zurich 29%, P<0.0001). There was a negative influence of age on visibility of the jejunal (OR=0.74, P<0.0001) and medial iliac lymph nodes (OR=0.86, P=0.010). Additionally, there was a positive influence of body weight on visibility of jejunal (OR=1.04, P=0.017) and medial iliac (OR=1.04, P=0.032) lymph nodes. Furthermore, gastrointestinal contents had a positive influence on visibility of the jejunal (OR=0.64, P=0.033) and medial iliac (OR=0.56, P=0.006) lymph nodes.
Discussion
In the present study, the pyloric area was visualized in 64% of dogs but the only factor having an effect on its identification was institution, with the pylorus being more frequently seen in Ghent. This may be due to the difference in patient positioning at the two institutes. Right recumbency in Zurich may have hampered the visibility of organs situated in the right abdomen because a table with an opening that allowed examination from the dependent side was not used.
The major duodenal papilla was visible in 42% of dogs. Its visibility was negatively influenced by gastrointestinal air and increased body weight. The negative impact of intestinal gas is not unexpected. Additionally, the cranial location of the major duodenal papilla in the abdomen also plays a role. At this level, it may be more difficult to exert sufficient pressure on the transducer to displace the small intestine and colon because of the ribs and the relatively fixed position of the transverse colon and duodenum. The body weight, rather than the body condition, score influenced visibility of the major duodenal papilla, which may also be related to the thoracic conformation.
The visibility of the pancreas was 60% for the body, 87% for the right lobe, and 56% for the left lobe. This corresponds to previous reports.8 Various factors prevent optimal sonographic examination of the pancreas, such as its ill-defined borders, respiratory motion, abdominal pain, and gastrointestinal gas.1,8,11 We confirmed the negative influence of gastrointestinal gas. Body weight, but not body condition score, also negatively influenced visualization of the pancreas. The influence of these parameters may also be related to the cranial location of the pancreas. The left pancreatic lobe was the only structure whose visualization was influenced positively by inherent image quality. In dogs, the left lobe is generally less visible than the right one, as it is situated in a region that may vary in appearance because of large collections of fat, splenomegaly, or stomachal or colonic contents.8 In our study, age did not have an influence on the visibility of the pancreas, in contrast to a previous report where the normal pancreas was more easily seen in young animals.1 Finally, the pancreas was seen more often in Zurich than in Ghent, probably related to the right recumbency in Zurich, which may eliminate the negative influence of overlying intestinal gas.
Ninety-one percent of the left and 86% of the right adrenal glands were visualized. In a previous study, the left adrenal gland was identified in 96% and the right in 72% of dogs.3 The right adrenal gland is more difficult to visualize than the left one due to its smaller size, deeper and more cranial location, respiratory motion, and the presence of gas in the pylorus, duodenum, and colon.3,5,12,13 Obesity and lack of patient compliance also interfere with the visualization of the adrenal glands.14 Here, the only factor influencing visualization of the adrenal glands was body condition score; a high body condition score had a positive influence on visibility of the right adrenal gland. It was unexpected to find a lack of an effect of body condition score on visualization of the organs studied herein, as obesity is often reported to result in bad image quality.1,3,12 A negative influence of obesity has been reported only for the right adrenal gland and not for the left one.2 This is contradictory to our study, where a high body condition score had a positive influence on the visibility of the right adrenal gland. The reasons for this finding are unclear. Fat surrounding the gland may increase the separation between the caudal vena cava and the right adrenal gland, resulting in a better definition of its margins. Also, there will be higher contrast between the hyperechoic fat and the slightly hypoechoic adrenal gland. Finally, a large body size or a deep thoracic conformation may interfere with visualization of the adrenal glands.3,12 However, we found no such negative influence.
The jejunal lymph nodes were detected in 51% of dogs and the medial iliac lymph nodes in 54%. This is lower than results described previously, with the right medial iliac lymph node seen in 82% and the left in 45% of normal dogs.7 The reason for this difference is unclear. Normal abdominal lymph nodes may not be detectable with ultrasonography.1,6,9 Lymph nodes were more frequently seen in Ghent than in Zurich. The different positioning of dogs may explain this discrepancy. The medial iliac lymph nodes are better imaged from a dorsolateral approach rather than a ventral one.7 With the dog in right recumbency, only the ventral approach can be used to detect the right medial iliac lymph node, and gas in the nondependent descending colon may interfere with detection of the left medial iliac lymph node. The reason why the jejunal lymph nodes were better seen in Ghent than in Zurich is unclear. In this study, there was a positive correlation between body weight and visibility of lymph nodes. This may be due to a direct relationship between lymph node size and body weight.6 Furthermore, lymph nodes were more frequently identified in young animals. This is due to the proportionally larger size of the lymph nodes in young compared with adult dogs.6,14 Finally, one unexpected result was the positive correlation between gastrointestinal content and lymph node visualization; this may be a spurious finding.
There was no effect of dog compliance on visualization of the organs of interest. The low number of incompliant dogs that were not sedated (17%) is probably the reason. However, in total, 28 dogs were incompliant, with 13 of those 28 dogs receiving sedation.
In conclusion, not all small abdominal organs will be found during a routine abdominal ultrasonographic examination. The normal duodenal papilla, pylorus, pancreatic body, left pancreatic lobe, and lymph nodes were only detected in 40–64% of the dogs. The organs that were most consistently visible (86–91%) were the right pancreatic lobe and both adrenal glands. The major factors influencing the visibility of the organs of interest in this study were body weight and gastrointestinal content. Age, body condition score, and inherent image quality were of only minor influence.
