DETERMINATION OF FISH GENDER USING FRACTAL ANALYSIS OF ULTRASOUND IMAGES

The gender of cod Gadus morhua can be determined by considering the complexity in their gonadal ultrasonographic appearance. The fractal dimension ( D _B) can be used to describe this feature in images. B-mode gonadal ultrasound images in 32 cod, where gender was known, were collected. Fractal analysis was performed on these images and D _B was determined using the box counting method. A receiver–operating curve (ROC) was drawn for D _B as a test for male fish. Using a range of D _B values, the maximum accuracy for this test was calculated and compared with the accuracy for identifying male fish by subjective analysis alone. The mean (and standard deviation) of the fractal dimension D _B for male fish was 1.554 (0.073) while for female fish it was 1.468 (0.061); the difference was statistically significant ( P =0.001). The area under the ROC curve was 0.84 indicating the value of fractal analysis in gender determination in cod. Maximum accuracy (0.84) for D _B as a test for male fish was obtained using the threshold value D _B=1.5058 compared with an accuracy of 0.78 for subjective image evaluation. The use of two thresholds, D _B<1.4475 (females) and D _B>1.5054 (males) gives an 80% certainty in the classification result. Fractal analysis is useful for gender determination in cod. This or a similar form of analysis may have wide application in veterinary imaging as a tool for quantification of complexity in images.     

ULTRASOUND APPEARANCE OF AN EQUINE TESTICULAR SEMINOMA

A 16-year-old thoroughbred stallion developed sudden swelling of the left testicle. The stallion had previously been regarded as a unilateral cryptorchid. Ultrasound examination of the left testicle revealed a diffusely heterogeneous parenchyma. The testicle was diffusely hypoechoic with ill defined regions of hyperechogenicity giving the appearance of hypoechoic nodules throughout the testicular parenchyma. No normal testicular tissue was identifiable. An echogenic band, representing a pseudocapsule could be seen surrounding the testicle. Histopathologic diagnosis a seminoma.     

CHARACTERIZATION OF CANINE SUPERFICIAL TUMORS USING GRAY-SCALE B MODE, COLOR FLOW MAPPING, AND SPECTRAL DOPPLER ULTRASONOGRAPHY—A MULTIVARIATE STUDY

Superficial tumors are not routinely evaluated by two- or three-dimensional diagnostic imaging methods as part of the staging of canine cancer patients, although superficial tumors are readily imaged by ultrasound. The objectives of this study were to characterize the ultrasonographic patterns of superficial tumors and to evaluate whether ultrasound can help discriminate between benign and malignant tumors in dogs. Superficial tumors (n=132) in 86 dogs were evaluated by B mode, color flow mapping, and spectral Doppler ultrasonography. Size, echogenicity, tumor border definition, invasiveness, acoustic transmission, presence and distribution of vascular flow to and within the tumor, as well as perfusion indices were measured. The tumors were classified as lipomas, benign tumors, atypical mammary tumors, and malignant tumors. Multivariate statistics using discriminant analysis was used to determine which parameters may be used to predict the status of the tumor. Tumor echogenicity, border shape, acoustic shadowing, total number of vessels to the tumor and the total flow amount are the parameters that in combination resulted in the lowest classification error (24%), meaning that on average three out of four tumors were correctly classified using these parameters. All the lipomas and atypical mammary tumors were classified correctly by ultrasonography. The results of this study show that ultrasonography has an important role in the evaluation of canine superficial tumors, particularly in the evaluation of tissue homogeneity and tumor vascularity.     

ESTIMATION OF FELINE RENAL VOLUME USING COMPUTED TOMOGRAPHY AND ULTRASOUND

Renal volume estimation is an important parameter for clinical evaluation of kidneys and research applications. A time efficient, repeatable, and accurate method for volume estimation is required. The purpose of this study was to describe the accuracy of ultrasound and computed tomography (CT) for estimating feline renal volume. Standardized ultrasound and CT scans were acquired for kidneys of 12 cadaver cats, in situ. Ultrasound and CT multiplanar reconstructions were used to record renal length measurements that were then used to calculate volume using the prolate ellipsoid formula for volume estimation. In addition, CT studies were reconstructed at 1 mm, 5 mm, and 1 cm, and transferred to a workstation where the renal volume was calculated using the voxel count method (hand drawn regions of interest). The reference standard kidney volume was then determined ex vivo using water displacement with the Archimedes’ principle. Ultrasound measurement of renal length accounted for approximately 87% of the variability in renal volume for the study population. The prolate ellipsoid formula exhibited proportional bias and underestimated renal volume by a median of 18.9%. Computed tomography volume estimates using the voxel count method with hand‐traced regions of interest provided the most accurate results, with increasing accuracy for smaller voxel sizes in grossly normal kidneys (–10.1 to 0.6%). Findings from this study supported the use of CT and the voxel count method for estimating feline renal volume in future clinical and research studies.     

EVALUATION OF EXPERIMENTALLY INDUCED CANINE HEPATIC CIRRHOSIS USING DUPLEX DOPPLER ULTRASOUND

Portal blood flow was measured with duplex Doppler ultrasound in ten normal dogs and in ten dogs with hepatic cirrhosis induced by common bile duct ligation 4 weeks previously. Mean portal blood flow velocity in the 10 dogs with experimentally induced hepatic cirrhosis was markedly reduced (9.2 ± 1.70 cm/sec vs. normal 18.1 ± 7.6 cm/sec, p < 0.01). Mean portal blood flow was also significantly decreased compared to normal (17.2 ± 4.9 cc/min/kg versus normal 31.06 ± 9.1 cc/min/kg, p < 0.01) while portal vein diameter remained unchanged. The dogs with induced hepatic cirrhosis developed extensive extrahepatic portosystemic shunting that was confirmed at necropsy. It was concluded that decreased portal velocity and portal flow which resulted from hepatic cirrhosis was detectable noninvasively with Doppler ultrasound.     

SONOGRAPHIC CHARACTERISTICS OF PRESUMPTIVELY NORMAL CANINE MEDIAL ILIAC AND SUPERFICIAL INGUINAL LYMPH NODES

The medial iliac and superficial inguinal lymph nodes are not routinely palpable in the dog, and ultrasound imaging provides an alternate noninvasive technique to assess these lymph nodes, as well as to guide needle aspiration. Herein we describe the ultrasound characteristics of the medial iliac and superficial inguinal lymph nodes in 50 healthy dogs, as well as frequency and ease of node detection. The relationship between the size of the lymph nodes and the following variables was assessed: age, gender, body weight, body condition score, body length, and thoracic height and width. Right and left medial iliac lymph nodes were detected in 50 (100%) dogs, right superficial inguinal lymph node(s) in 49 (98%) dogs, and left superficial inguinal lymph node(s) in 47 (96%) dogs. In >90% of both sets of lymph nodes, the echogenicity was hypoechoic or isoechoic to surrounding tissues, with a corticomedullary or homogenous echotexture, smooth, clearly defined margins, and a fusiform shape. Increasing weight, distance from the sternal manubrium to the ischium, and thoracic height and width were associated with increased lymph node size (P‐values<0.05). Average lymph node sizes and range of sizes provide preliminary reference values for the medial iliac and superficial inguinal lymph nodes in normal dogs.     

EVALUATION OF THE ABDOMEN USING ULTRASOUND FOLLOWING A VENTRAL MIDLINE CELIOTOMY

Over the last 20 years, ultrasound has become a useful diagnostic tool in veterinary medicine. The purpose of this study was to use ultrasonographic examinations to determine if intra-abdominal contents could be visualized. Nineteen dogs, ranging from 8 months to 18 years of age were included. Each animal was evaluated sonographically prior to surgery, 24 h after surgery, and then 2-6 weeks following the surgical procedure. The results of our study suggest that ultrasound is an effective means for evaluation of the abdomen and surgery site following surgery. Both the abdomen and surgery site were clearly visualized in all dogs. In addition, subjective evaluation of the intra-abdominal contents was performed. The limiting factor in this study was not the presence of air, as was expected. Rather the limiting factor for appropriate visualization was pain. The information obtained could prove to be useful to surgeons and radiologists in assessing a patient's status postoperatively.     

ULTRASOUND IMAGING OF THE COELOMIC STRUCTURES IN THE BOA CONSTRICTOR (BOA CONSTRICTOR)

Transverse ultrasound images from three live boa constrictors ( Boa constrictor ) were compared to the corresponding gross sections of a euthanized boa constrictor. The snakes were anesthetized, placed into dorsal recumbency, and imaged through the ventral scutes. Ultrasound images were produced with a real time, B-mode scanner and a linear 7.5-MHz transducer. The heart, liver, stomach, gall bladder, pancreas, kidneys, and fat body were the coelomic structures most consistently recognized ultrasonographically. Results of this study indicate that ultrasound is a useful imaging technique in snakes.     

QUANTITATIVE EVALUATION OF IMAGENT® AS AN ABDOMINAL ULTRASOUND CONTRAST MEDIUM IN DOGS

Diagnostic ultrasound currently provides a noninvasive, economical means to evaluate internal organ structure. Methods to increase the amount of diagnostic information obtained from the sonographic examination are being explored. A new area of research has recently focused on the ability of microbubble contrast agents to help provide additional information. A prospective study was undertaken to quantitatively evaluate the amount of gray scale enhancement of ultrasound images of six canine livers, spleens, and left kidneys that occurs after intravenous injection of Imagent. Computerized analysis was used to assess the percent change in brightness of these images. A transient decrease in the percent change in brightness was seen to occur after injection of higher doses of Imagent. This effect was more pronounced in the deeper portions of the organs of interest. Imagent provides a measurable change in the brightness of the organs of interest and would be potentially useful in a clinical setting.     

CHARACTERIZATION OF LYMPHOMATOUS LYMPH NODES IN DOGS USING CONTRAST HARMONIC AND POWER DOPPLER ULTRASOUND

Doppler ultrasound has been used in humans to determine angioarchitecture of lymph nodes as the criterion for the determination of malignancy. We hypothesized that the vascular and perfusion patterns of a canine malignant lymph node could be characterized with intravenous microbubble ultrasound contrast media and that contrast harmonic ultrasound could provide better conspicuity of the angioarchitecture when compared with Power Doppler ultrasound. In this study, 11 peripheral lymph nodes in dogs with histologically verified malignant lymphoma were imaged with fundamental ultrasound, Power Doppler ultrasound, and three contrast harmonic pulse sequences to characterize the vascular pattern and perfusion. Vascular imaging was greatly enhanced in these nodes with 2.13 times more vessels seen with contrast harmonic ultrasound compared with Power Doppler ultrasound (P < 0.01). The angioarchitecture of lymphomatous lymph nodes of dogs in this study were similar to those previously described in malignant superficial lymph nodes in human patients; 45.5% of the nodes had displacement of the central hilar vessel, 45.5% had aberrant vessels, 63.6% had pericapsular vessels, 36.4% had subcapsular vessels, and 81.8% had loss of the central hyperechoic band in fundamental sonography. Poor perfusion, indicated by a lower mean pixel intensity increase between pre- and postcontrast administration images, was seen in 36.4% of the lymph nodes while 63.6% had fair to good perfusion. The perfusion patterns in nine of the 11 lymph nodes were homogenous and two showed focal hypoperfused regions. We conclude that Power Doppler and contrast harmonic ultrasound are beneficial in accurately depicting angioarchitechture and can provide additional information in determining the presence of malignant vascular characteristics within lymphomatous nodes in dogs.     

IMAGING DIAGNOSIS—ULTRASOUND UNCOMMON FEATURES OF AN ABDOMINAL GOSSYPIBOMA IN A DOG

An abdominal mass was incidentally detected in a 12‐year‐old, neutered female, crossed breed dog. Abdominal ultrasonographic examination showed a well‐delineated, irregular, heterogeneous mass that did not generate any distal acoustic shadowing. Transcutaneous US‐guided biopsy of the mass were nonconclusive but raised the possibility of neoplasia. Surgery discovered a mass embedded in the omentum and a large quantity of surgical sponges were identified in cut section. To the authors’ knowledge, this represents the first published case of gossypiboma casting no characteristic distal acoustic shadowing.     

ENDOSCOPIC ULTRASOUND OF THE CANINE ABDOMEN

The aim of this study was to develop a standardized procedure for examination of the canine abdomen using endoscopic ultrasound and to describe the organs and structures that could be identified transgastrically. The abdomen of four healthy dogs and two cadavers were examined with endoscopic ultrasound. Five anatomic landmarks were used for standardized imaging of the cranial abdomen. These were the portal vein, splenic head and body, duodenum, left kidney, and aorta. High-resolution images of the following organs and structures could be made: distal esophagus, gastric wall from the cardia to the pylorus, liver, caudal vena cava, hepatic lymph nodes, liver hilus, and associated vessels, trifurcation of the celiac artery as well as the path of its branches and the left pancreatic limb and body. Structures that were more difficult to image were the distal duodenum and right pancreatic limb, the entire jejunum, ileum, and cecum as well as the tail of the spleen. Endoscopic ultrasound allowed excellent visualization of the gastric wall and regional structures without interference with gas artefacts. Centrally located organs such as the pancreas could be well examined transgastrically with endoscopic ultrasound without interference by overlying intestinal segments as is common with transabdominal ultrasound.     

CHARACTERIZATION OF CANINE FOCAL LIVER LESIONS WITH CONTRAST-ENHANCED ULTRASOUND USING A NOVEL CONTRAST AGENT—SONAZOID

Contrast-enhanced ultrasound using Sonazoid, a novel contrast medium with a liver-specific Kupffer phase, was evaluated in canine focal liver lesions Twenty-five dogs with a liver mass were given intravenous Sonazoid, and the enhancement pattern in the arterial, portal, and parenchymal phase was characterized. An enhancement defect in the lesion in the parenchymal phase was observed in all malignant lesions, whereas only one of nine benign lesions had a filling defect. The diagnostic value of the presence of a filling defect for malignancy was statistically significant (100% sensitivity, 88.9% specificity, 94.1% positive predictive value, 100% negative predictive value), and was equal to that of hypoenhancement in the portal or delayed phase. The defect pattern (clear or irregular defect) was dependent ( P <0.05) on the types of malignancy (i.e., hepatocellular carcinoma and other types of malignancies). In the arterial phase, five of the six hepatocellular carcinomas had hypervascularity, whereas no other lesion was characterized by hypervascularity. In some dogs, additional lesions that could not be observed with conventional B-mode ultrasonography were detected in the parenchymal phase. The enhancement pattern of Sonazoid, especially in the parenchymal phase, has potential as a diagnostic tool for canine focal liver lesions.     

MEASUREMENTS OF HINDLIMB BLOOD FLOW RECORDED USING DOPPLER ULTRASOUND DURING ADMINISTRATION OF VASOACTIVE AGENTS IN HALOTHANE-ANESTHETIZED HORSES

The purpose of the study was to determine the ability of Doppler ultrasound to detect changes in femoral blood flow during pharmacologic manipulation of arterial blood pressure. Doppler ultrasonography was performed in the femoral vessels of six halothane-anesthetized horses before and during administration of phenylephrine HCI and sodium nitroprusside. The time-averaged mean velocity and volumetric flow were calculated. The contour of the velocity waveform was assessed, and the early diastolic deceleration slope (EDDS) and pulsatility index (PI) were calculated. Administration of phenylephrine HCI resulted in increased mean aortic blood pressure (MABP) by 40% (29.3-53.0%). This caused significant decrease in cardiac output (26.8 to 13.5 l/min), femoral arterial velocity (left artery 7.20 to 4.00 cm/s; right artery 5.01 to 3.39 cm/s) and volumetric flow (left artery 556 to 221 ml/min; right artery 397 to 193 ml/min) in the femoral vessels and significant increase in systemic vascular resistance (163 to 433 dyn-s/cm5), EDDS (1a: 285 to 468: ra: 250 to 481) and PI (1a: 9.38 to 20.4; ra 17.1 to 29.1). Administration of sodium nitroprusside resulted in a decreased MABP of 27.2% (22.5-33%). This increased cardiac output (20.8 to 32.4 L/min), however, no significant changes were observed in femoral blood flow. Despite obvious changes in the waveform contour, no significant change occurred in EDDS or PI. These results suggest that Doppler ultrasound may be useful for measuring femoral blood flow in anesthetized horses. However, waveform analysis appears to be limited when multiple changes occur in central and peripheral haemodynamics.     

ULTRASOUND PREDICTION OF PARTURITION IN QUEENS

Eight pregnancies in five queens were followed with serial ultrasound examinations. Measurements of head and body diameter were made throughout pregnancy. These measurements were used to produce individual growth curves for each pregnancy and average curves for both head and body diameter. Formulas based on the slope of the growth curves were derived. These formulas were then used to predict parturition dates based on single measurements of fetal head and/or body diameter in eight additional queens. Predicted date of parturition was then compared to actual queening dates. Ultrasound prediction of parturition date was accurate to within two days in seven of eight queens.     

QUANTITATIVE CONTRAST HARMONIC ULTRASOUND IMAGING OF NORMAL CANINE LIVER

Eight adult dogs with no evidence of liver disease, weighing between 8 and 25 kg were imaged after injection of a microbubble contrast medium using harmonic ultrasound imaging. All dogs received three separate bolus contrast injections, and six dogs also received three separate constant rate infusions each. Time/Mean Pixel Value curves were generated for selected regions of the liver. Upslope, downslope, baseline, peak, change, and time to peak were calculated. For bolus injection (averaging all subjects), upslope was 3.85 +/- 1.50 Mean Pixel Values/s, downslope was -0.71 +/- 0.30 Mean Pixel Values/s, baseline was 72.38 +/- 17.82 Mean Pixel Values, peak was 120.26 +/- 17.44 Mean Pixel Value, change from baseline to peak was 47.88 +/- 6.92 Mean Pixel Values, time to peak (from injection) was 22.88 +/- 6.79 s, and time to peak (from first upslope) was 13.88 +/- 1.55 s. Data acquisition and analysis from constant rate infusions was more cumbersome than for bolus, and results were less repeatable. There were significant differences (p < .005) in upslope, downslope, peak values, and time to peak between the two methods. These baseline data may prove useful in the evaluation of dogs with diffuse hepatic disease.     

ULTRASOUND EVALUATION OF SMALL BOWEL THICKNESS COMPARED TO WEIGHT IN NORMAL DOGS

The use of ultrasound to measure small bowel thickness is an important part of any ultrasound examination of the abdomen. Increased thickness of the intestinal wall is a hallmark for the detection of diseases ranging from inflammatory bowel disease to neoplasia. Our subjective impression has been that dogs with no clinical signs of gastrointestinal disease often have sonographic measurements greater than published norms. The purpose of this study was to prospectively reevaluate these norms. The clinical history on all dogs receiving an abdominal ultrasound examination was evaluated for signs of gastrointestinal disease. Those without clinical signs were entered into this study. The documentation of body weight, breed, jejunal thickness, and duodenal thickness was made in 231 dogs. Dogs were placed into five groups based on their weight. Sixty-nine breeds were represented with weight ranging from 2.1 to 64 kg. A statistically significant (P < or = 0.05) correlation between body weight and both jejunal and duodenal thickness was observed. We also found that the maximum thickness in both jejunum and duodenum in healthy dogs was larger than previously reported. These data indicate norms for the jejunum of < or = 4.1 mm for dogs up to 20 kg, < or = 4.4 mm for dogs between 20 and 39.9 kg, and < or = 4.7 mm for dogs over 40 kg. The data indicate norms for the duodenum < or = 5.1 mm for dogs up to 20 kg, < or = 5.3 kg for dogs between 20 and 29.9 kg, and < or = 6.0 mm for dogs over 30 kg.     

ULTRASOUND EVALUATION OF THE COXOFEMORAL JOINTS OF THE CANINE NEONATE

The effectiveness of ultrasound in depicting the normal anatomy of the coxofemoral joint in puppies was evaluated. Ultrasound images were obtained from twenty Springer Spaniel puppies ranging from one day to twelve weeks of age. The anatomy of the hip joint could be identified up to eight weeks of age. By twelve weeks of age the degree of ossification prevented complete hip joint evaluation. Three different views were utilized. The longitudinal and dorsolateral views were the most useful. Motion of the femoral head within the acetabulum could be observed. Measurements of the acetabulum, femoral head and joint space could be obtained however these measurements were inconsistent. The technique appears to be useful for depicting the anatomy of the neonatal coxofemoral joint however, additional studies utilizing both normal and dysplastic puppies are recommended to determine if this examination will be useful in discriminating between normal and dysplastic puppies.     

THE ROLE OF ULTRASOUND IN THE ASSESSMENT OF LARYNGEAL PARALYSIS IN THE DOG

Forty dogs with clinical signs suggestive of upper respiratory tract disease underwent echolaryngography and laryngoscopy. Laryngoscopy was used as the definitive technique to diagnose laryngeal paralysis. The ultrasound investigation accurately indicated the presence of the paralysis and confirmed the uni- or bilateral nature of the disorder. Findings indicative of laryngeal paralysis included asymmetry or absence of motion of the cuneiform processes (30/30), abnormal arytenoid movement (16/30), paradoxical movement (9/30), caudal displacement of the larynx (2/30) and laryngeal collapse (1/30). Thirty dogs were found to be afflicted with laryngeal paralysis and ten had normal laryngeal motility.     

QUANTITATIVE CONTRAST ULTRASOUND ANALYSIS OF RENAL PERFUSION IN NORMAL DOGS

Eight normal dogs with no evidence of renal disease, weighing between 8 and 25 kg were imaged using contrast harmonic ultrasound after injection of a microbubble contrast medium. All dogs received three separate bolus injections of 0.05 ml of commercial contrast medium (Definity). Time/mean pixel value (MPV) curves were generated for selected regions in the cortex and medulla of the left kidney in each dog. Upslope, downslope, baseline, peak intensity, and time to peak were calculated for each zone. For a bolus injection, within the renal cortex (averaging all subjects) the upslope was 7.4 +/- 1.5 MPV/s, downslope was -0.4 +/- .2 MPV/s, baseline was 66.8 +/- 9.3 MPV, peak was 103.6 +/- 8.2 MPV, time to peak (from injection) was 12.8 +/- 5.3 s and from time of contrast medium reaching the kidney was 5.1 +/- 2.0 s. Within the renal medulla (averaging all subjects), upslope was 2.8 +/- 1.7 MPV/s, downslope was -0.3 +/- .2 MPV/s, baseline was 39.3 +/- 6.0 MPV, peak was 65.2 +/- 14.3 MPV, time to peak from injection was 20.9 +/- 6.4 s and from time of contrast reaching the kidney was 11.6 +/- 4.1 s. These baseline data may prove useful in the evaluation of dogs with diffuse disease or vascular compromise.