Shannon P. Holmes, MSc, DVM
Mention of a vomiting dog or cat immediately evokes a relatively large list of differential diagnoses. Clinical evaluation and analysis of blood may eliminate some, but diagnostic imaging ultimately plays a major role in the achievement of a definitive diagnosis. With regard to diagnostic imaging, the first-line approach can vary greatly between different veterinary hospitals. Depending on clinician experience or preference, some perform abdominal radiography and some perform abdominal ultrasonography as their first diagnostic step. The doctors reviewing abdominal radiographs or performing abdominal ultrasound can range in experience from a relatively inexperienced intern to internal medicine or surgery specialists to radiologists.
Abdominal radiography is not routinely performed in human medicine due to increased usage of abdominal computed tomography (CT). The sequela of this decrease in abdominal radiographic procedures is decreased frequency of abdominal radiographic interpretation and, by extension, loss of interpretative skills. A similar pattern is seen in veterinary medicine, where abdominal ultrasound is replacing abdominal radiography as a first line diagnostic tool. This practice appears to be related to practitioners having greater confidence in ultrasound over their radiographic interpretative skills. The downside of this practice is that abdominal ultrasound lacks the sensitivity and specificity possessed by CT and therefore cannot replace it in the same manner. The new twist is actually an old practice revisited. In any veterinary abdominal exam, but especially in vomiting patients, the standard of practice should be the institution of radiography prior to all ultrasound exams, since the two imaging modalities represent complimentary diagnostic tests.
The ease of performing abdominal radiography and sharing the radiographs has been greatly improved by digital radiography (DR) and teleradiology, respectively. DR is poised to impact the veterinary community in much the same manner as ultrasound. The American College of Veterinary Radiology (ACVR) has released a supplemental publication dedicated to the many facets of DR and teleradiology, in recognition of the importance of this new technology to the veterinary community. The reality of DR in clinical practice is the ability to rapidly obtain high quality orthogonal radiographs of the abdomen within a few minutes. An ultrasound examination of the entire abdomen requires more time, the quality is greatly influenced by operator skill level and care must be taken to thoroughly evaluate all structures. The simplicity and speed of DR with or without the use of teleradiology should re-popularize abdominal radiography as the first-line diagnostic imaging modality.
Case examples demonstrating the benefit of abdominal radiography as a first-line diagnostic tool will be discussed. This session will include facets of digital radiography as they relate to abdominal radiographic interpretation, with special focus on vomiting patients.
Abdominal Radiography--As a First-Line Diagnostic Approach
A discussion of abdominal radiographs should always utilize descriptors of Roentgen findings. Abdominal interpretation requires a systematic assessment of the opacity, size, shape, number and location of organs. In the time constraints of daily practice, this process is often detrimentally abbreviated. A contributor to this process is the bias every clinician possesses due to knowledge gained from the history, physical examination and blood analyses. Radiographic interpretation should be the same regardless of this information; it is the radiographic conclusions and differential diagnoses that are influenced.
While there are only 5 radiopacities, critical information is often gleaned from radiographs that can be overlooked on ultrasound or even preclude the need for ultrasound. Anything that increases the contrast on abdominal radiographs is typically going to involve a disease process that may be less defined on abdominal ultrasound. The most classic example is free abdominal gas. This is easily overlooked on ultrasound, as every sonographers' instinct is to apply pressure on the abdomen via the transducer to improve transmission of the ultrasound beam. This action displaces the gas away from the transducer and it will go undetected. Intra-abdominal fat also improves radiographic contrast. On ultrasound, it increases reverberation artifact, thereby reducing transmission of the ultrasound beam into deep abdominal structures. Conversely, free fluid reduces abdominal radiographic contrast and in these cases, ultrasound can outperform radiography. Occasionally peritonitis will reduce radiographic contrast and produce significant sonographic reverberation artifact compromising assessment of abdominal organs. These demonstrate the complimentary nature of the two modalities. The important point to recognize is that there is no reliable way to predict the contrast of an abdominal radiograph and thus, it should remain a useful first-step in the clinical course of a diagnostic work-up.
A radiograph is a two dimensional representation of a three dimensional structures, but it contains important spatial data that can be difficult to accurately assess on ultrasound. Interpretation of mass effect on abdominal radiographs can provide crucial information regarding potential organ(s) involvement. Without this information prior to abdominal ultrasound, the displacement of abdominal organs can disorient a sonographer which in turn compromises the quality of the study. An entire radiograph can quickly be reviewed, but a complete abdominal ultrasound is time consuming. It is more common to perform targeted ultrasound examination based on history, physical exam findings and blood work. This risk in this practice is missing lesions or other abnormalities that may influence clinical decisions. Abdominal radiography and ultrasonography are complimentary in that radiographs display spatial relationships of abdominal organs and ultrasound display cross-sectional information about the abdominal organs. In this manner, the two modalities can provide information similar to CT and the need for both components of information is the reason that CT has replaced this practice in human patients.
Finally, the extra-abdominal structures can be assessed with abdominal radiographs. Vomiting associated with caudal esophageal disease is often noted on abdominal radiographs. Causes of vomiting, such as hiatal herniation, can only be diagnosed with radiography, since the ultrasound beam cannot penetrate into the thorax.
Abdominal Digital Radiography
The introduction and application of DR in a practice requires the not only the adaptation to interpreting radiographs on monitors, but a thorough understanding of how the images are acquired and processed. In this manuscript, DR is referring to images that are presented in a digital format. It is beyond the scope of this manuscript to cover all aspects of DR. Digital images can be produced by either computed radiography (indirect method) or direct digital radiography. These differ in the process of image acquisition, but the end product is an image presented in a digital format that is typically viewed on a computer monitor (sometimes called soft copy images).
One major reported benefit of DR is a reduction in retakes due to incorrect exposure. A wider range of exposure techniques will produce a diagnostic abdominal radiograph. However, the same exposure can produce radiographs differing in appearance or enhancing visualization of different structures. For example, it is possible to re-process an abdominal radiograph to produce radiographs that enhance the spinal column. The end product of DR is a function of the look-up tables (LUT). These tables assign a specific brightness levels to the digital signal received. It is the application of different LUTs that determine the appearance of a digital radiograph and these can be selected by the operator.
One of the most common errors in digital radiograph production is the processing of an abdominal digital radiograph using a thoracic digital radiograph LUT. This is an example of technician error that DR cannot correct for. If the applied LUT has too narrow a range of values, some pixel data will be lost. This is referred to clipping. Radiographically, portions of structures that are known to be present are not seen on the radiograph. In addition, this information cannot be retrieved with re-windowing the image. When the appropriate LUT is used, each structure in the image is seen as if the technique had been set specifically for that structure. In large deep chested dogs, correct exposure of the caudal abdomen on regular film-screen radiographs will often require a second radiograph because the cranial abdomen is underexposed. With DR and the correct LUT and preprocessing, the entire abdomen is uniformly visualized. In addition, the details of the caudal thorax and musculoskeletal structures should be seen with the same detail.
In regular film-screen radiography, underexposure results in the image being too light. In DR, underexposure produces a uniform grainy or pixilated appearance in the image. In making digital radiographs, the exposure number needs to be within a specific range to produce a diagnostic radiograph. It is often possible to correct for incorrect exposure numbers, except in the case of extreme underexposure and overexposure. In these respective scenarios, insufficient information is acquired by the DR plate to form an image, or the DR plate becomes over saturated and cannot make the image.
Other artifacts encountered occur during exposure. Because of the high sensitivity of the DR plates to radiation, any non-uniformity in the radiation beam can contribute to the image. Each system is calibrated for the x-ray tube and any structures overlying the plate (collimator, table top, etc.). These non-uniformities are removed from the image in the processing step. With miscalibration or a change in the imaging system, silhouettes of external structures can artifactually produce lesions. Another exposure artifact is ghost images. This occurs when information from a prior radiograph is readout during the digitization of the later radiograph. It most commonly occurs when digital radiographs are made in quick succession.
These are the most common artifacts encountered, but other DR artifacts will be encountered some of which of an undetermined etiology or related to proprietary information. The artifacts of regular film-screen radiographs that everyone has learned to read through and correct are different than those experienced with DR. CR will produce some that are similar to regular film-screen radiographs, but also can produce artifacts unique to this type of acquisition. Artifact recognition requires practice and very thorough radiographic interpretation.
Similar to physician radiology, there are not sufficient numbers of board certified veterinary radiologists in the United States to supply all hospitals. With teleradiology, there are now veterinary radiologists that are available 24 hours a day, seven days a week. If needed, interpretation can be rendered within 20-30 minutes of submission. The immediacy of these interpretations can impact cases in hospitals around the country. Access to radiographic interpretation by board certified radiologists can alleviate the some of the anxiety that is commonly associated with abdominal radiographic interpretation. This interaction also aids in the development of practitioners' radiographic interpretive skills.
The best format for transmitting radiographs to a radiologist is DICOM 3.0 standard. DICOM is the most direct method for transmission of radiographs. In physician radiology, interpretation of images that are non-DICOM format, for example JPEG format, is not possible due to legal issues. The ACVR is in the process of reviewing DICOM standard, which is expected to impact the industry and specifically those vendors marketing non-DICOM format digital images. At this time, the decision to interpret non-DICOM format images as part of their teleradiology services is at the discretion of the individual veterinary radiologists.
Until CT is ubiquitously available to the veterinary community as a diagnostic imaging tool in vomiting patients, abdominal radiography should remain the first-line diagnostic modality. It is a quick, global assessment of the abdomen and can dictate the next diagnostic step. In a 2006 study, a group of physician doctors, including radiologists, internist, surgeons and criticalists, had their ability to interpret plain abdominal radiographs tested. The results of study showed that specialized training and experience were the major contributing factors influencing interpretation quality. Good abdominal film interpretation requires practice. This translates into a need for increased use of abdominal radiography and retrospective analysis of radiographs with knowledge gained from subsequent ultrasound examination or surgical findings to improve interpretative skills. If the veterinary community does not continue to use and develop is radiographic interpretative skills, they will be lost.
1. Bushberg JT. The essential physics of medical imaging. (2nd Ed.) Lippincott, Williams & Wilkins (2001).
2. Petite A, et al. Radiographic and ultrasonographic findings of emphysematous cystitis in four nondiabetic female dogs. Vet Radiol Ultrasound 2006; 47: 90.
3. Hoffman KL. Sonographic signs of gastroduodenal linear foreign bodies in 3 dogs. Vet Radiol Ultrasound 2003; 44: 466.
4. Boysen SR, et al. Ultrasonographic findings in dogs and cats with gastrointestinal perforation. Vet Radiol Ultrasound 2003; 44: 556.
5. Mazzaferro EM. Triage approach to the acute abdomen. Clin Tech Small Anim Pract 2003; 18: 1.
6. Poteet BA. Veterinary Teleradiology. Vet Radiol Ultrasound Supplement Volume 49 (1); S33.
7. Drost WT, et al. Digital radiography artifacts. Vet Radiol Ultrasound Supplement 2007; 49: S48
8. Mattoon JS. Digital radiography. Vet Comp Orthop Traumatol 2006; 19: 123.
9. Lim C, et al. Plain abdominal radiographs: Can we interpret them? Ann R Coll Surg Engl2006; 88: 23.
10. Thompson W. Accuracy of Abdominal Radiography in Acute Small-Bowel Obstruction: Does Reviewer Experience Matter? AJR:188, March 2007; W233.
11. Boone J. Radiological interpretation 2020: toward quantitative image assessment. Med Phys. 2007 Nov;34(11):4173.