Abstract
OBJECTIVE To determine anatomic reference points for 4 turtle species and to evaluate data on relative anatomic dimensions, signal intensities (SIs), and position of selected organs within the coelomic cavity by use of MRI.
ANIMALS 3 turtle cadavers (1 red-eared slider [Trachemys scripta elegans], 1 yellow-bellied slider [Trachemys scripta scripta], and 1 Coastal plain cooter [Pseudemys concinna floridana]) and 63 live adult turtles (30 red-eared sliders, 20 yellow-bellied sliders, 5 Coastal plain cooters, and 8 hieroglyphic river cooters [Pseudemys concinna hieroglyphica]).
PROCEDURES MRI and necropsy were performed on the 3 turtle cadavers. Physical examination, hematologic evaluation, and whole-body radiography were performed on the 63 live turtles. Turtles were sedated, and MRI in transverse, sagittal, and dorsal planes was used to measure organ dimensions, position within the coelomic cavity, and SIs. Body positioning after sedation was standardized with the head, neck, limbs, and tail positioned in maximum extension.
RESULTS Measurements of the heart, liver, gallbladder, and kidneys in sagittal, transverse, and dorsal planes; relative position of those organs within the coelom; and SIs of the kidneys and liver were obtained with MRI and provided anatomic data for these 4 turtle species.
CONCLUSIONS AND CLINICAL RELEVANCE MRI was a valuable tool for determining the position, dimensions, and SIs of selected organs. Measurement of organs in freshwater chelonians was achievable with MRI. Further studies are needed to establish reference values for anatomic structures in turtles. Results reported here may serve as guidelines and aid in clinical interpretation of MRI images for these 4 species.
Diagnostic imaging plays an important role in chelonian medicine because of the limitations of physical examination in this vertebrate order.1 Because of the unique anatomy of chelonians, including the shell, adequate palpation of internal structures is extremely difficult.2 Radiographic techniques are insufficient for detailed visualization of most of the internal organs in turtles and tortoises.3,4 Although ultrasonography is a valuable tool for examining soft tissues in many other reptile species such as snakes5 and lizards,6–8 it is often of restricted value for use on small chelonians.9,10 Visual examination of internal structures such as the gastrointestinal tract, urogenital tract (including the testes, ovaries, follicles, urinary bladder, and kidneys), heart, liver, and gallbladder is possible but limited to the inguinal, prefemoral, axillary, and cervical regions available as acoustic windows in turtles and tortoises, especially for smaller species.3,11 Because of the surrounding shell and resulting small windows for an ultrasonographic probe,12 visible examination of soft tissues within the coelomic cavity in small chelonians, such as most pet tortoises and fresh water turtles, is restricted to parts of organs. Consequently, it is hard or impossible to measure the entire dimensions of organs ultrasonographically. Computed tomography has been used to examine the respiratory tract13 and bone tissues,14 but it has limits for other soft tissues. Particularly for chelonians, CT 3-D and multiplanar reconstruction might serve as a valuable tool to illustrate normal anatomic structures15 and pathological conditions.16
Magnetic resonance imaging is an excellent tool for evaluation of the dimensions and positions of parenchymatous organs within the coelomic cavity despite the surrounding shell.12,14,17 Examination of organs via MRI has been reported for various tortoise and turtle species.18,19 Use of MRI for chelonians consists of case reports20–22 of a single animal or studies23–26 on relatively small numbers of turtles and tortoises. Magnetic resonance imaging investigations in larger numbers of animals have been conducted in Europe on turtles27 and tortoises.28 Several investigations in chelonians have focused on the comparison of cross-sectional anatomy of the animals with MRI scans.29–32 Intracoelomic anatomy in sea turtles has been investigated by use of MRI.29,30 In 1 study,33 investigators used CT and described changes in the dimensions of the lungs attributable to body positioning and extension of the neck and extremities in 14 red-eared sliders (Trachemys scripta elegans).
The objective of the study reported here was to obtain anatomic reference points by use of MRI for 4 species of North American turtles commonly kept as pets and to collect additional data on the anatomic dimensions of organs of these turtle species, the position of organs within the coelomic cavity, and the SIs. This information could serve as a basis for future routine investigations on measurements and evaluation of dimensions and locations of organs within the shell of chelonians.
Materials and Methods
Animals
Cadavers of 3 turtles that died of trauma (1 red-eared slider, 1 yellow-bellied slider [Trachemys scripta scripta], and 1 Coastal plain cooter [Pseudemys concinna floridana]) were examined radiographically. Necropsy of each turtle followed immediately thereafter.
Subsequently, 63 healthy adult turtles (36 females and 27 males), consisting of 4 North American species commonly kept by individual households and zoological parks in Germany, were examined. The turtles comprised 30 red-eared sliders, 20 yellow-bellied sliders, 5 Coastal plain cooters, and 8 hieroglyphic river cooters (Pseudemys concinna hieroglyphica). The turtles were a part of various collections in Lower Saxony, Germany. Written consent was obtained from clients and zoological parks for use of all turtles. Use of animals for this study was approved by the Institutional Animal Care and Use Committee of the University of Veterinary Medicine Hanover, Germany.
Examination of cadavers
The 3 turtle cadavers were examined radiographically with a digital systema and by use of MRI with an applied MRI unit.b Necropsy was performed immediately thereafter and included identification and measurement of various organs. The purpose was to verify the identity of organs (including the heart, lungs, liver, gallbladder, and kidneys) and other structures of the gastrointestinal and urogenital tracts and compare them with those identified for images obtained in the 3 planes of the MRI. In this manner, MRI images in various planes were used to examine organ morphology, location, and size for sliders and cooters and subsequently serve as guidelines for the measurement of organ dimensions by use of MRI in live chelonians of these and related species.
Examination of live turtles
A total of 63 healthy adult turtles were examined. Body weight ranged between 300 and 3,500 g. Turtles were housed outdoors during the summer. All turtles were housed indoors during the winter, and none of them hibernated. Thus, even during the winter months, all of them remained active and continued to eat. The diet consisted of omnivore nourishment, such as commercially available turtle sticks, Gammarus spp, lettuce, and fish. At least 24 to 48 hours before turtles were sedated, they were transported to our university veterinary clinic and housed at room temperature (24° to 26°C) in 200-L open tanks filled with approximately 100 L of water. Each tank provided turtles with access to a small area with a dry surface. At least 8 hours before MRI was performed, turtles were moved onto wet towels located in the same tanks.
Physical examination of each turtle was performed, and body weight was obtained. Length, width, and height of the carapace of each turtle were measured by use of a Vernier caliper and recorded. Carapace length was measured on the midline from the cranial edge of the nuchal scute to the caudal edge of the caudalmost marginal scutes. Carapace height was measured as the most dorsal point (typically at the carapace vertebra 3) to the plastron os abdominale. Carapace width was measured as the greatest distance between the anterior and posterior bridges at the height of vertebrae 2 and 3.
A blood sample (1 mL) was collected from the dorsal tail vein. The PCV and 19 hematologic biochemical variables (activity of alanine aminotransferase, glutamate dehydrogenase, alkaline phosphatase, aspartate aminotransferase, cholinesterase, and creatine kinase and concentrations of BUN, sodium, potassium, total calcium, ionized calcium, phosphorus, fructosamine, cholesterol, glucose, uric acid, albumin, total protein, and total bilirubin) were measured for each turtle. Animals with severe changes in these hematologic variables were excluded from the study.
Digital radiographs were obtained in dorsoventral, lateral, and craniocaudal projections by use of a digital system.a Data were stored on an optical disk; data were analyzed and digitized with a data-processing unit.c Radiographs were photographed with a laser camerad that featured an image-developing unit and stored on hard drives. Radiographs were used to detect possible (radiopaque) metallic foreign bodies within the turtles, which was a criterion for exclusion from the study. Other criteria for exclusion were clinical and radiographic signs of pneumonia or other severe pathological changes within the coelomic cavity.
Sedation
Approximately 30 minutes before MRI was performed, turtles were sedated by injection of medetomidinee (0.2 to 0.3 mg/kg) and ketamine hydrochloridef (10 to 15 mg/kg) into the musculature of the right and left forelimbs. This allowed consistent stress-free positioning of animals during the subsequent MRI procedures. Sedation was monitored by use of the corneal reflex and withdrawal reflex and evaluation of muscle relaxation. After MRI was completed, atipamezoleg (0.75 mg/kg) was injected into the musculature of a forelimb.
MRI
Magnetic resonance imaging was performed as whole-body scans and included 3 projections (transverse, sagittal, and dorsal), each with T1- and T2-weighted images. The MRI unitb had a gantry bore of 58 cm in diameter and a superconducting magnet (cooled with helium) from a niobium-titanium alloy, which produced a field strength of 1.0 T. During MRI, an extremity coil was used for smaller turtles and a head coil was used for larger turtles. These coils were designed for evaluation of limbs and the head of humans and allowed exact positioning of the turtles. For all figures, the settings for T1-weighted MRI were as follows: repetition time = 330 milliseconds and the echo time = 12 milliseconds. Settings for window width and level were adjusted to obtain optimal images. Calculated data were exported with an image-editing programh in DICOM data format and archived on compact disks.
Sedated turtles were placed in ventral recumbency and craniocaudal and laterolateral alignment so that the plastron rested horizontally on the pad. Warmed gel pads were used to assist in correct positioning and provide warmth to the sedated turtles during MRI. During the study, extreme care was taken to ensure that body position after sedation was the same for all turtles. Therefore, the head, neck, limbs, and tail of each animal were positioned in maximum extension and affixed with tape to the MRI table throughout all measurements.
Turtles were scanned in toto in the transverse, dorsal, and sagittal planes, each with T1- and T2-weighted protocols. Depending on the size of the turtle, 3- or 4-mm slice cross-sectional images were obtained. Smaller cross-sectional images (3-mm slices) were chosen for smaller turtles because of the smaller size of structures and organs. To ensure reproducibility of the measurements, the plane in which the organ had the largest dimension or the plane that included the view of a particular anatomic structure that could be clearly defined in every turtle was selected for the measurements. Because of differences in intensity within the organs, we did not measure SIs by use of individual points for MRI. Instead, the largest possible ellipsoidal areas were selected with the intention of obtaining more generalized values (including arithmetic mean values) for organs.
MRI measurements
The T1-weighted images were evaluated in dorsoventral, laterolateral transverse, and sagittal planes. The plane or planes used for measurements were selected on the basis of anatomic landmarks (which were exactly defined in each turtle) or, to ensure reproducibility of the measurements, in the plane in which the largest dimension of the organ was measured. Furthermore, to avoid error, SIs were measured in the largest ellipsoidal areas, and arithmetic means were calculated from these data rather than by obtaining individual focal points.
Results of the study were based on selected patterns of 20,967 individual MRI images. Because of the size variation of the turtles and the goal of obtaining a representative cross section of the population, statistical analysis of relative organ dimensions, rather than absolute organ dimensions, was performed. To obtain relative organ dimensions, a second representative value for a turtle (eg, carapace height or carapace length) was obtained, and variables for measured organ size were calculated and reported as ratios.
MRI transverse T1-weighted images
Heart—Cardiac variables were measured in the transverse plane. The heart had the greatest dorsoventral and laterolateral dimensions in this plane.
For the dorsoventral dimension of the heart, a line that served as the first baseline parallel to the widest dimension of the plastron was drawn (Figure 1; line 1). Subsequently, a line was drawn parallel to the plastron at the height of the most ventral aspect of the heart and served as a second baseline (line 2). An additional parallel line was drawn at the most dorsal aspect of the heart (line 3). The distance between lines 2 and 3 was measured; the ratio of that distance and carapace height, defined in the same manner as the ventral aspect of the plastron (in the broadest possible dimension) and the most dorsal aspect of the carapace, was calculated.
Transverse T1-weighted MRI images of a male yellow-bellied slider (Trachemys scripta scripta) used to determine the dorsoventral dimension of the heart (A) and the relative dorsoventral position of the heart within the coelomic cavity (B). In panel A, a line that served as the first baseline parallel to the widest dimension of the plastron was drawn (line 1). Subsequently, a line was drawn parallel to the plastron at the height of the most ventral aspect of the heart and served as a second baseline (line 2). An additional parallel line was drawn at the most dorsal aspect of the heart (line 3). The distance between lines 2 and 3 was measured. In panel B, an additional line was drawn at the most dorsal point of the carapace (line 4), and the distance between lines 3 and 4 was measured. Settings for window width and window level were 3,085 and 1,178 HU, respectively. AS = Ventral. L = Left. PI = Dorsal. R = Right.
Citation: American Journal of Veterinary Research 78, 12; 10.2460/ajvr.78.12.1387
The relative dorsoventral position of the heart within the coelomic cavity was determined (Figure 1). To obtain the distance between the dorsal aspect of the heart and carapace height, an additional line was drawn at the most dorsal point of the carapace (line 4). The distance between lines 3 and 4 was measured; the ratio of that distance and carapace height was calculated as the relative dorsoventral heart position within the coelomic cavity.
The laterolateral dimension of the heart was determined in the same transverse MRI image as was used to determine the dorsoventral dimensions of the heart (Figure 2). A line that served as the first baseline parallel to the widest dimension of the plastron was drawn (line 1). Two parallel lines were drawn perpendicular to line 1 at the widest dimension of the heart at the most lateral limits on the right (line 2) and left (line 3) side. The distance between lines 2 and 3 was measured.
Transverse T1-weighted MRI images of a male yellow-bellied slider used to determine the laterolateral dimension of the heart (A) and the relative laterolateral position of the heart within the coelomic cavity (B). In panel A, a line that served as the first baseline parallel to the widest dimension of the plastron was drawn (line 1). Two parallel lines were drawn perpendicular to line 1 at the widest dimension of the heart at the most lateral limits on the right (line 2) and left (line 3) side. The distance between lines 2 and 3 was measured. In panel B, an additional line parallel to the lines at the lateral limits of the heart was drawn at the outermost points of the lateral aspect of the right side of the carapace (line 4); this line was used for measurements used to calculate ratios. Settings for window width and window level were 3,085 and 1,178 HU, respectively. AS = Ventral. L = Left. PI = Dorsal. R = Right.
Citation: American Journal of Veterinary Research 78, 12; 10.2460/ajvr.78.12.1387
To obtain the relative laterolateral position of the heart within the coelomic cavity, additional lines parallel to the ones at the lateral limits of the heart were drawn at the outermost points of the lateral aspect of the left and right sides of the carapace; these lines were perpendicular to line 1 (Figure 2). The ratios between the laterolateral position of the heart and the lateral aspect of the carapace to the right and between the laterolateral position of the heart and the lateral aspect of the carapace to the left were calculated.
Kidneys—Width and height of the right and left kidneys were determined in the transverse plane. Sectional images were selected in which the width and height of both kidneys were the greatest.
For the measurement of the dorsoventral dimension (height) of the kidneys, a line that served as the first baseline parallel to the widest dimension of the plastron line was drawn (Figure 3; line 1). Subsequently, lines were drawn parallel to line 1 at the height of the most ventral aspect of the right and left kidneys (line 2). Lines also were drawn at the most dorsal aspect of the right and left kidneys (line 3). The distance between lines 2 and 3 was measured for each kidney and used as the value for the dorsoventral height of the right and left kidneys. Ratios were calculated and defined in the same manner as the ventral aspect of the plastron (in the widest possible dimension) and the most dorsal aspect of the carapace.
Transverse T1-weighted MRI images of a male yellow-bellied slider used to determine the dorsoventral (A) and laterolateral (B) dimensions of the kidneys. In panel A, a line that served as the first baseline parallel to the widest dimension of the plastron was drawn (line 1). Subsequently, lines were drawn parallel to line 1 at the height of the most ventral aspect of the right and left kidneys (line 2) and the most dorsal aspect of the right and left kidneys (line 3). The distance between lines 2 and 3 was measured for each kidney and used as the value for the dorsoventral height of the right and left kidneys. In panel B, 2 additional lines (line 4 and 5) that were perpendicular to line 1 were drawn at the most lateral and most medial limits of the right kidney. The distance between lines 4 and 5 was measured and was the width of the right kidney. Settings for window width and window level were 970 and 768 HU, respectively. AS = Ventral. L = Left. P = Dorsal. R = Right.
Citation: American Journal of Veterinary Research 78, 12; 10.2460/ajvr.78.12.1387
The laterolateral dimension (width) of the right and left kidneys was also determined in the transverse image (Figure 3). The distance was measured between 2 parallel lines (line 4 and 5) that were perpendicular to line 1 and at the most lateral and most medial limits of each kidney. The distance between lines 4 and 5 was measured and was the width of each kidney. To determine the relative laterolateral position of both kidneys within the coelomic cavity, additional lines were drawn at the outermost points of the lateral aspect of the left and right sides of the carapace and used to calculate a ratio.
SI of the kidneys and musculature
Mean SI of each kidney (Figure 4) for the largest ellipsoidal area possible was determined in an attempt to minimize measurement errors. Mean SI of the retractor capitis muscle served as a reference value for the evaluation. A circular area (4 mm in diameter) of the retractor capitis muscle located between the 2 kidneys was used as a reference tissue. Measured values were reported, and a ratio between the SI values was calculated.
Transverse T1-weighted MRI image of a male yellow-bellied slider used to determine the SI of the kidneys and SI of the retractor capitis muscle. The largest ellipsoidal area possible for each kidney (ks) was used to minimize measurement errors of SI. A circular area (4 mm in diameter) located centrally in the area of the retractor capitis muscle between the 2 kidneys (ms) was used to determine the muscle SI. Settings for window width and window level were 1,446 and 820 HU, respectively. AS = Ventral. L = Left. P = Dorsal. R = Right.
Citation: American Journal of Veterinary Research 78, 12; 10.2460/ajvr.78.12.1387
MRI sagittal T1-weighted images
Kidneys—To determine the extent of each kidney in the sagittal plane, images were chosen in which each kidney had the greatest dimensions (Figure 5). A line was drawn connecting the most dorsal and most ventral points at the most cranial aspect of each kidney (line 1). Additional lines were drawn to determine the dorsoventral position of each kidney within the coelomic cavity in relation to the carapace height. To obtain the distance between the most ventral aspect of each kidney and the plastron, lines were drawn at the most ventral point of the plastron (line 2) and at the most ventral point of each kidney (line 3). Measurements between these lines were determined, and ratios were calculated between those values and carapace height.
Sagittal T1-weighted MRI image of a male yellow-bellied slider used to determine the dorsoventral dimensions of the kidneys and distance to the outer limits of the ventral aspect of the plastron. A dashed line was drawn connecting the most dorsal and most ventral points at the most cranial aspect of each kidney (line 1). Additional lines were drawn at the most ventral point of the plastron (line 2) and at the most ventral point of each kidney (line 3). Settings for window width and window level were 1,843 and 769 HU, respectively. AS = Ventral. I = Caudal. P = Dorsal. S = Cranial.
Citation: American Journal of Veterinary Research 78, 12; 10.2460/ajvr.78.12.1387
Dimensions of the gallbladder and right lobe of the liver in the craniocaudal direction
Dimensions for the gallbladder were determined in the sagittal plane because the gallbladder had the largest dimension in that plane (Figure 6). A line was drawn connecting the most dorsal point to the most ventral point of the gallbladder (line 1). A line was drawn at the ventral aspect of the plastron and parallel to the widest dimension of the plastron (line 2). The craniocaudal dimension of the right lobe of the liver was determined in the same sagittal image. The distance was measured between 2 parallel lines that were perpendicular to line 2 and located at the most cranial (line 3) and most caudal (line 4) limits of the right lobe of the liver. These measurements were used to determine the largest dimensions of the gallbladder and to calculate a ratio in relation to the craniocaudal dimension of the right lobe of the liver.
Sagittal T1-weighted MRI image of a male yellow-bellied slider used to determine the craniocaudal dimension of the right lobe of the liver and the dimensions of the gallbladder. A dashed line was drawn connecting the most dorsal point to the most ventral point of the gallbladder (line 1). A line was drawn at the ventral aspect of the plastron and parallel to the widest dimension of the plastron (line 2). The distance between 2 parallel lines that were perpendicular to line 2 and located at the most cranial (line 3) and most caudal (line 4) limits of the right lobe of the liver was measured. Settings for window width and window level were 1,683 and 681 HU, respectively. AS = Ventral. I = Caudal. P = Dorsal. S = Cranial.
Citation: American Journal of Veterinary Research 78, 12; 10.2460/ajvr.78.12.1387
MRI dorsal (coronal) T1-weighted images
Dimensions of the right lobe of the liver in the craniocaudal direction—The dorsal plane was used as the reference image to determine craniocaudal dimensions of the right lobe of the liver because it was the plane in which the right lobe of the liver had the largest dimensions (Figure 7). A line was drawn centrally through the tuber of the right and left lateral acromion processes (line 1). This line served as a guideline for 2 parallel lines that were perpendicular to line 1 and drawn at the most cranial (line 2) and most caudal (line 3) points of the right lobe of the liver. The distance between lines 2 and 3 was measured and used as the value for the craniocaudal dimensions of the right lobe of the liver, and a ratio between that value and the measured carapace length was calculated.
Dorsal T1-weighted MRI image of a male yellow-bellied slider used to determine the craniocaudal dimension of the right lobe of the liver. A line was drawn centrally through the tuber of the right and left lateral acromion processes (line 1). Two parallel lines that were perpendicular to line 1 were drawn at the most cranial (line 2) and most caudal (line 3) points of the right lobe of the liver. The distance between lines 2 and 3 was measured and used as the value for the craniocaudal dimension of the right lobe of the liver. Settings for window width and window level were 2,666 and 611 HU, respectively. I = Caudal. L = Left. R = Right. S = Cranial.
Citation: American Journal of Veterinary Research 78, 12; 10.2460/ajvr.78.12.1387
SI of the right lobe of the liver and musculature
The measured value of the mean SI of the liver was determined from the largest ellipsoidal area of the right lobe of the liver cranial to the gallbladder (Figure 8). The use of the large ellipsoidal area was intended to minimize measurement errors. In addition, the arithmetic mean of a defined area of the muscles of the shoulder joint served as a reference value. The arithmetic mean of a 4-mm-diameter area of the shoulder joint musculature (deltoid muscle) located laterally from the right scapula (which served as an anatomic landmark) was determined. Measured values for each tissue were reported, and the ratio of the values was calculated.
Dorsal T1-weighted MRI image of a female hieroglyphic river cooter (Pseudemys concinna hieroglyphica) used to determine the SI of the right lobe of the liver and the SI of the musculature of the shoulder joint. The largest ellipsoidal area possible for the right lobe of the liver cranial to the gallbladder (ls) was used to minimize measurement errors of SI. A circular area (4 mm in diameter) located laterally from the right scapula in the deltoid muscle (ms) was used to determine the muscle SI. Settings for window width and window level were 3,166 and 645 HU, respectively. I = Caudal. L = Left. R = Right. S = Cranial.
Citation: American Journal of Veterinary Research 78, 12; 10.2460/ajvr.78.12.1387
Statistical analysis
To eliminate the possible influence of body size on organ-size comparisons among turtle species, organ dimensions were adjusted by dividing the value by the appropriate body size measurements. Resulting values were reported as ratios in the analyses.
For all measured variables, the arithmetic mean, SD, and minimum and maximum values were calculated. Data were assessed for normal distribution by use of the Shapiro-Wilk test and visual assessment of Q-Q plots of model residuals. Differences of organ-size measurements among the 4 species were analyzed by use of a 1-way ANOVA34,i and post hoc Ryan-Einot-Gabriel-Welsch multiple range test for multiple pairwise comparisons.35 Organ-size measurements of paired organs were compared with a t test for paired observations within species.
For the assessment of reliability of measurements, 7 variables of various SIs in 12 animals were each measured 3 times. The ICC describes the relationship between the variance among animals and the total variance among animals and within animals for several repetitions of measurements. Reliability is high when the effect of repeated measurements is small. The ICC was calculatedj by variance component estimation as follows:
where VCanimal is the variance component for animal, and VCrepeated measures is the variance component for repeated measures. For all analyses, values were considered significant at P < 0.05.
Results
Examination of cadavers
Results of MRI and subsequent necropsy of a red-eared slider, yellow-bellied slider, and Coastal plain cooter were used to verify anatomic structures identified in the MRI images. Organs that included the heart, liver, gallbladder, and kidneys as well as muscles and bones evident in 3 planes of the MRI images corresponded well to the anatomic structures of all turtles examined during necropsy.
Examination of live turtles
Measurements were obtained for all 63 turtles of the 4 species (Table 1). All turtles were successfully sedated with medetomidine and ketamine and underwent MRI without any movement. All turtles were capable of swimming within 1 hour after IM injection of atipamezole.
Characteristics of red-eared sliders (Trachemys scripta elegans), yellow-bellied sliders (Trachemys scripta scripta), Coastal plain cooters (Pseudemys concinna floridana), and hieroglyphic river cooters (Pseudemys concinna hieroglyphica) examined with MRI to obtain measurements of structures in the coelomic cavity.
| Species and sex | n | Body weight (g) | Carapace length (mm) | Carapace width (mm) | Carapace height (mm) |
|---|---|---|---|---|---|
| Red-eared slider | |||||
| Female | 17 | 510–1,880 | 141–225 | 112–169 | 61–96 |
| Male | 13 | 340–778 | 126–166 | 93–133 | 47–66 |
| Yellow-bellied slider | |||||
| Female | 9 | 710–2,000 | 173–220 | 131–174 | 64–102 |
| Male | 11 | 380–725 | 133–162 | 107–125 | 57–69 |
| Coastal plain cooter | |||||
| Female | 2 | 1,440–2,596 | 200–255 | 144–180 | 90–118 |
| Male | 3 | 791–1,174 | 169–198 | 129–150 | 73–76 |
| Hieroglyphic river cooter | |||||
| Female | 8 | 656–3,515 | 145–296 | 102–203 | 64–106 |
| Male | 0 | NA | NA | NA | NA |
Values reported represent minimum to maximum. Carapace length was measured on the midline from the cranial edge of the nuchal scute to the caudal edge of the caudalmost marginal scutes. Carapace height was measured as the most dorsal point (typically at the carapace vertebra 3) to the plastron os abdominale. Carapace width was measured as the greatest distance between the anterior and posterior bridges at the height of vertebrae 2 and 3.
NA = Not applicable.
No abnormalities were detected in results of physical examination, hematologic analysis, and digital radiography. Therefore, all of the turtles were considered healthy and included in the study. Use of MRI provided excellent visualization of all organs and structures in all 63 turtles. The T1-weighted MRI images provided the best visibility of the anatomic structures (eg, heart, kidneys, liver, gallbladder, and musculature) in transverse, sagittal, and dorsal planes. Thus, all measurements of the organ positions within the coelomic cavity, the organ dimensions, and the SIs were performed on T1-weighted images. Results of MRI investigations were summarized for transverse (Table 2) and sagittal and dorsal (Table 3) planes.
Arithmetic mean ± SD values for turtles examined by use of MRI in the transverse plane.
| Variable | Red-eared slider (n = 30) | Yellow-bellied slider (n = 20) | Coastal plain cooter (n = 5) | Hieroglyphic river cooter (n = 8) |
|---|---|---|---|---|
| Ratio of dorsoventral diameter of the heart to carapace height | 0.32 ± 0.04a,b | 0.29 ± 0.04b | 0.30 ± 0.01b | 0.34 ± 0.04a |
| Ratio of dorsoventral position of the heart to carapace height | 0.57 ± 0.04 | 0.59 ± 0.06 | 0.55 ± 0.03 | 0.56 ± 0.03 |
| Ratio of laterolateral diameter of the heart to carapace width | 0.26 ± 0.03a,b | 0.24 ± 0.03b | 0.25 ± 0.04b | 0.29 ± 0.06a |
| Ratio of the laterolateral position of the heart to the lateral aspect of the carapace to the right | 0.34 ± 0.03 | 0.35 ± 0.03 | 0.34 ± 0.01 | 0.35 ± 0.07 |
| Ratio of the laterolateral position of the heart to the lateral aspect of the carapace to the left | 0.36 ± 0.02 | 0.36 ± 0.03 | 0.36 ± 0.02 | 0.37 ± 0.06 |
| Ratio of dorsoventral diameter of the right kidney to carapace height | 0.31 ± 0.07 | 0.32 ± 0.07 | 0.24 ± 0.06 | 0.26 ± 0.05 |
| Ratio of dorsoventral diameter of the left kidney to carapace height | 0.31 ± 0.07 | 0.25 ± 0.06 | 0.25 ± 0.06 | 0.24 ± 0.04 |
| Ratio of laterolateral diameter of the right kidney to carapace width | 0.09 ± 0.02 | 0.10 ± 0.01 | 0.09 ± 0.01 | 0.10 ± 0.02 |
| Ratio of laterolateral diameter of the left kidney to carapace width | 0.09 ± 0.02 | 0.10 ± 0.01 | 0.09 ± 0.01 | 0.10 ± 0.02 |
| Ratio of SI of the right kidney to SI of the retractor capitis muscle | 1.24 ± 0.09 | 1.27 ± 0.13 | 1.28 ± 0.13 | 1.15 ± 0.16 |
| Ratio of SI of the left kidney to SI of the retractor capitis muscle | 1.27 ± 0.10 | 1.30 ± 0.12 | 1.34 ± 0.13 | 1.22 ± 0.17 |
Values with different superscript letters differ significantly (P < 0.05; Ryan-Einot-Gabriel-Welsch multiple range test).
Arithmetic mean ± SD values for turtles examined by use of MRI in the sagittal and dorsal planes.
| Plane | Variable | Red-eared slider (n = 30) | Yellow-bellied slider (n = 20) | Coastal plain cooter (n = 5) | Hieroglyphic river cooter (n = 8) |
|---|---|---|---|---|---|
| Sagittal | Ratio of dorsoventral position of the right kidney to carapace height | 0.46 ± 0.05 | 0.42 ± 0.06 | 0.43 ± 0.05 | 0.44 ± 0.03 |
| Ratio of dorsoventral position of the left kidney to carapace height | 0.45 ± 0.04 | 0.43 ± 0.07 | 0.41 ± 0.05 | 0.45 ± 0.03 | |
| Ratio of dorsoventral position of the right kidney to carapace length | 0.16 ± 0.03 | 0.16 ± 0.03 | 0.17 ± 0.03 | 0.17 ± 0.03 | |
| Ratio of dorsoventral position of the left kidney to carapace length | 0.16 ± 0.03 | 0.16 ± 0.03 | 0.17 ± 0.03 | 0.17 ± 0.03 | |
| Ratio of dimensions of the gallbladder to craniocaudal extent of the liver | 0.39 ± 0.13 | 0.34 ± 0.07 | 0.31 ± 0.11 | 0.33 ± 0.06 | |
| Dorsal | Ratio of craniocaudal dimension of the liver to carapace length | 0.34 ± 0.06 | 0.37 ± 0.05 | 0.36 ± 0.05 | 0.37 ± 0.10 |
| Ratio of SI of the liver to the SI of the shoulder joint musculature | 2.01 ± 0.47 | 2.02 ± 0.36 | 1.90 ± 0.15 | 1.97 ± 0.44 |
MRI transverse T1-weighted images
Position of the heart and heart dimensions in dorsoventral and laterolateral directions—Relative dorsoventral diameter of the heart (MRI transverse plane) was determined and calculated in relation to the carapace height (Table 2). Minimum and maximum values were 0.21 and 0.40, respectively; there were significant differences among the species. Relative dorsoventral position of the heart within the coelomic cavity was determined and calculated in relation to the carapace height. Minimum and maximum values were 0.39 and 0.67, respectively, and no significant species-specific differences were found. Relative laterolateral diameter of the heart was determined and calculated in relation to the carapace width. Minimum and maximum values were 0.18 and 0.44, respectively; there were significant differences among the species.
Relative laterolateral position of the heart within the coelomic cavity was determined. Therefore, the ratio between the laterolateral position of the heart and the lateral aspect of the carapace to the right was calculated (Table 2). Minimum and maximum values were 0.24 and 0.52, respectively, and no significant species-specific differences were found. Relative laterolateral position of the heart within the coelomic cavity was determined. The ratio between the laterolateral position of the heart and the lateral aspect of the carapace to the left was calculated. Minimum and maximum values were 0.28 and 0.51, respectively, and no significant species-specific differences were found. The distance from the left side of the heart to the lateral aspect of the carapace to the left was significantly (P < 0.01) greater than the distance from the right side of the heart to the lateral aspect of the carapace to the right.
Position of the kidneys and their dimensions in dorsoventral and laterolateral directions—Relative dorsoventral height of the right kidney (MRI transverse plane) was determined and calculated in relation to the carapace height (Table 2). Minimum and maximum values were 0.17 and 0.47, respectively, and no significant species-specific differences were found. Relative dorsoventral height of the left kidney was determined and calculated in relation to the carapace height. Minimum and maximum values were 0.18 and 0.47, respectively, and no significant species-specific differences were found. There were no significant differences in the dorsoventral height between the left and right kidneys.
Relative laterolateral width of the right kidney was determined and calculated in relation to the carapace width (Table 2). Minimum and maximum values were 0.03 and 0.14, respectively, and no significant species-specific differences were found. Relative laterolateral width of the left kidney was determined and calculated in relation to the carapace width. Minimum and maximum values were 0.06 and 0.14, respectively, and no significant species-specific differences were found. There were no significant differences in the laterolateral width of the left and right kidneys.
SI of the kidneys and musculature—The SI of the right kidney and retractor capitis muscle were measured (MRI transverse plane) and indicated as a ratio for the right kidney (Table 2). Minimum and maximum values were 0.87 and 1.53, respectively, and no significant species-specific differences were found. The SI of the left kidney and capitis retractor muscle were also measured and calculated as a ratio for the left kidney. Minimum and maximum values were 0.91 and 1.58, respectively, and no significant species-specific differences were found. The SI of the left kidney differed significantly (P < 0.01) from that of the right kidney.
MRI sagittal T1-weighted images
Positions and dimensions of the kidneys in a dorsoventral direction in the sagittal plane—Relative position of the right kidney within the coelomic cavity (MRI sagittal plane) was determined and calculated in relation to the carapace height (Table 3). Minimum and maximum values were 0.30 and 0.57, respectively, and no significant differences were detected among the species. Relative position of the left kidney within the coelomic cavity was determined and calculated in relation to the carapace height. Minimum and maximum values were 0.30 and 0.62, respectively, and no significant species-specific differences were found.
Relative dimensions of the right kidney were determined and calculated in relation to the carapace length (Table 3). Minimum and maximum values were 0.09 and 0.26, respectively, and no significant species-specific differences were found. Relative dimensions of the left kidney were determined and calculated in relation to the carapace length. Minimum and maximum values were 0.06 and 0.25, respectively, and no significant species-specific differences were found. The left kidney was not significantly (P = 0.079) longer than the right kidney.
Dimensions of the liver in the craniocaudal direction and dimensions of the gallbladder in the sagittal plane—Relative size of the liver in the craniocaudal direction and dimensions of the gallbladder (ie, maximum diameter of the gallbladder in relation to the craniocaudal extent of the liver measured on the same sagittal image) were determined, and gallbladder dimensions were calculated in relation to the craniocaudal extent of the liver (Table 3). Minimum and maximum values were 0.21 and 0.73, respectively, and no significant species-specific differences were found.
MRI dorsal T1-weighted images
Dimensions of the liver in the craniocaudal direction in the dorsal plane—Relative craniocaudal dimensions of the liver (MRI dorsal plane) were determined and calculated in relation to the carapace length (Table 3). Minimum and maximum values were 0.21 and 0.60, respectively, and no significant species-specific differences were found.
SI of the liver and shoulder joint musculature in the dorsal plane—The SI of the liver and SI of the shoulder joint musculature were measured (MRI dorsal plane) and calculated as a ratio (Table 3). Minimum and maximum values were 1.17 and 3.00, respectively, and no significant species-specific differences were found.
Variance component analysis
Variance component analysis was conducted by 3-fold repetition of 7 measurement variables of various SIs of 12 turtles. The SIs were determined as a calculation of an arithmetic mean for multiple points, and the ICC for animals and measurement errors were calculated for the assessment of reproducibility (Table 4). Values of ρ ranged from 0.98 to 0.99 (ie, only 1% to 2% of variation depended on repeated measurements), which indicated extremely high reproducibility for the method used.
Results of variance component analysis for triplicate measurements of SIs of turtles by use of MRI in each of 3 planes.
| Component | Variance component (%) | ||||
|---|---|---|---|---|---|
| Plane | Variable | Error | Animal | Repeated measurement | Animal |
| Transverse | SI of right kidney | 112.72 | 11,272.2 | 0.99 | 99.01 |
| SI of left kidney | 121.01 | 12,101.0 | 0.99 | 99.01 | |
| SI of retractor capitis muscle | 3.41 | 4,835.5 | 0.07 | 99.93 | |
| Sagittal | SI of right gluteal muscles | 6.13 | 4,687.7 | 0.13 | 99.87 |
| SI of left gluteal muscles | 13.58 | 5,040.2 | 0.27 | 99.73 | |
| Dorsal | SI of liver | 2.30 | 24,934.6 | 0.01 | 99.99 |
| SI of shoulder joint musculature | 51.15 | 2,200.6 | 2.27 | 97.73 | |
Discussion
Results of the study reported here should serve as guidelines for measuring organ dimensions and their anatomic positions in the coelomic cavity. These results should help to establish values for the heart, kidneys, liver, gallbladder, and SIs of various intracoelomic organs in 4 North American species of semiaquatic freshwater turtles.
For the cadaver examinations, organ identification and size measurements on MRI images were compared with anatomic structures seen during whole-body necropsy. Structures including the heart, liver, gallbladder, kidneys, and muscles identified by means of MRI images were crossmatched and verified by use of cadaver gross anatomy. The cadaver examinations confirmed that MRI images provided excellent representations of the internal anatomic structures of sliders and cooters. Measurement of organ dimensions obtained by use of MRI corresponded with direct measurement of anatomic structures of turtles obtained during necropsy. The MRI images of the head and intracoelomic organs of various species of chelonians have been placed alongside frozen anatomic sections for comparison and, in some cases, for the creation of atlases of cross-sectional MRI anatomy of sea turtles, such as green turtles (Chelonia mydas) and loggerhead sea turtles (Caretta caretta)29,30,32; tortoises, especially Hermann tortoises (Testudo hermanni)18,28; and freshwater turtles, such as red-eared sliders27,31 and yellow-bellied sliders.31
None of the turtles in the present study hibernated. This avoided possible changes of organ size that were described in a study28 in Germany of Hermann tortoises in which pulmonary dimensions increased and stomach size decreased during hibernation.
Animals in the present study were anesthetized and placed in ventral recumbency with the head, neck, and all extremities fully extended during MRI. This is in contrast to other studies in which unsedated tortoises were partially immobilized on special wooden blocks1,18,27,31,36 or foam blocks37 or were taped in position with a fabric hook-and-loop fastener26 for MRI. In some investigations, most of the tortoises or turtles underwent MRI by simply covering the head opening or all shell openings with medical tape19 or covering all shell openings except the head opening.28 Some researchers have used hypothermia (ambient temperature between 6° and 9°C) for immobilization of tortoises just before or after hibernation.28 In most of the previous studies,1,18–26,28,30,31,36,37 turtles were placed in ventral recumbency for MRI. Investigators of a recent study33 of CT in red-eared sliders found that placing turtles in various recumbent positions (horizontal ventral, vertical left lateral, right lateral, and caudal recumbent positions) significantly affected lung volume measurements. This led some authors to conclude that positioning could also potentially affect interpretation of radiographs,33 especially for intracoelomic organs. Authors of 1 study17 stated that chelonian patients should be positioned in ventral recumbency for horizontal radiographic beams because coelomic contents gravitate ventrally. Some clinicians believe that positioning turtles in dorsal recumbency is stressful and may displace the organs, which would make interpretation difficult. Also, tipping a patient may cause abnormal displacement of internal organs and create radiographic artifacts.36
Positioning of the head and limbs is essential for obtaining good-quality radiographic images of chelonians and also appears to impact intracoelomic organ positioning. Some authors have stated that retraction of the head, neck, and limbs into the shell during radiographic examination results in superimposition of the limb musculature on the internal organs and coelomic viscera.12,38 Moreover, radiopacity and radiographic details of coelomic contents are affected by the position of the extremities in chelonians.17 The position of the head, neck, and extremities can play a role for other imaging modalities such as CT or MRI as well. In CT scans of red-eared sliders, variations in position of the head, neck, and extremities resulted in differences in lung volume as a result of space occupation during withdrawal into the shell.33 These findings led to our assumption that organ positions and sizes may change during MRI as a result of changes in the positioning of the animal as well as that of the head, neck, and extremities. Therefore, we believe it is essential to have the head and limbs of chelonians extended during all imaging procedures. Because most turtles and tortoises will retract the head, neck, and limbs to protect themselves, which leads to changes of organ positions and sizes, sedation or anesthesia is considered essential for standardized correct positioning of chelonian patients, especially for time-consuming imaging diagnostic procedures such as CT or MRI.
The selected sedation protocol (medetomidine and ketamine) used in the present study allowed MRI to be performed on all turtles in the required positions without any interference. All turtles regained coordinated swimming ability by 1 hour after administration of the antagonist atipamezole. Several authors have conducted studies to evaluate the effects of ketamine, xylazine, and midazolam39 or medetomidine-ketamine40 in red-eared sliders. Various sedation and chemical restraint protocols have been described for noninvasive MRI diagnostic evaluation of chelonians.18,22,27,29,30,36 For repeated CT evaluations of 14 red-eared sliders, dexmedetomidine (0.1 mg/kg), midazolam (1.0 mg/kg), and ketamine (2.0 mg/kg) were injected SC 45 to 60 minutes prior to CT, and turtles received flumazenil and atipamezole SC to reverse effects after completion of CT scans.33 The medetomidine, ketamine, and atipamezole protocol used in the study reported here enabled us to obtain good MRI results without detrimental effects to the turtles.
For the study reported here, MRI provided excellent soft tissue contrast and yielded highly detailed images of most soft tissue structures, as has been described previously.12,14,18,41 The MRI examinations of North American freshwater turtles in the present study were performed with a 1.0-T MRI machine with extremity or head coils comprising series of 3- or 4-mm cross-sectional images and that included transverse, dorsal, and sagittal planes with T1- and T2-weighted protocols to obtain good-quality MRI images for every turtle. In other investigations of chelonians, MRI units with 0.5-T,19,20,26 1.0-T,1,27,31 or 1.5-T1,21–24,28–30,37 magnets have been used with good results. The MRI examinations of larger turtles (eg, green turtles and loggerhead sea turtles29,30) have been performed without coils. Use of MRI in studies of smaller turtles and tortoises have included the use of knee coils,20,22,37 surface coils,27,31 or head coils.1,18,26,28 Slice thickness for MRI has differed among studies, chelonians, and planes (1.5 to 4 mm,28 3 to 5 mm,19 3 to 6 mm,27,31 4 to 5 mm,22 4 to 6.6 mm,30 7 mm,21 and 4, 5 to 7, or 10 mm29). In the study reported here, the use of extremity or head coils and a slice thickness of 3 or 4 mm led to good results for all 4 North American species of freshwater turtles.
For MRI in the present study, we used T1- and T2-weighted protocols, which have been used in most studies1,18–20,22,26–31,42 of chelonians, with varying results. In the study reported here, T1-weighted images had the best results for measurements of most structures (including the heart, liver, gallbladder, kidneys, and muscle). To facilitate and decrease the time needed to obtain MRI images for organ measurements, we used T1-weighted images for all organ size, position, and SI measurements. Some researchers have used both T1- and T2-weighted images of sea turtles,29,30 Hermann tortoises,28 red-eared sliders,27,31 or yellow-bellied sliders31 to obtain the best results for MRI evaluations. Authors of 1 study26 preferred T1-weighted images for MRI to evaluate the gastrointestinal tract of Hermann tortoises. Authors of another study31 of sliders found that T2-weighted images provided more detail for most of the structures, except for ovarian follicles and fat tissue. In a study19 of a variety of chelonian species, liver parenchyma was best evaluated on T1-weighted images, and fluid and the gallbladder were better visualized on T2-weighted images. Therefore, it appears most suitable to use both T1- and T2-weighted images to detect subtle changes within structures, as has been suggested elsewhere.42 For the present study, we obtained both T1- and T2-weighted images and then chose the T1-weighted results for all measurements to limit the number of MRI images used for evaluations.
Results of the present study indicated that imaging and measurement values of the dimensions and positions of the heart, liver, gallbladder, and kidneys as well as relative SIs of the liver, musculature, and kidneys can be reliably obtained by use of MRI for adult sedated red-eared sliders, yellow-bellied sliders, Coastal plain cooters, and hieroglyphic river cooters. Values were reported as ratios to provide a valid assessment with regard to the difference in sizes of the turtles between sexes and among species. The high reproducibility of the results, as determined on the basis of variance component analysis of various SIs, indicated that the measurements obtained in the present study can be reliably repeated.
Acknowledgments
This manuscript represents a portion of a thesis submitted by Dr. Schnack to the Clinic for Small Mammals, Reptiles and Birds at the University of Veterinary Medicine Hannover as partial fulfillment of the requirements for a Doctor Medicinae Veterinariae degree.
The authors declare that there were no conflicts of interest.
Presented in abstract form at the 14th Annual Conference of the Association of Reptilian and Amphibian Veterinarians, New Orleans, April 2007; the 17th Annual Conference of the Association of Reptilian and Amphibian Veterinarians, South Padre Island, Texas, October 2010; and the 21st Annual Conference of the Association of Reptilian and Amphibian Veterinarians, Portland, Ore, August 2016.
The authors thank Elliott Jacobson for technical review of the manuscript.
ABBREVIATIONS
| HU | Hounsfield units |
| ICC | Intraclass correlation coefficient |
| SI | Signal intensity |
Footnotes
Bucky diagnost, OPTIMUS 50, Philips Medical Systems, Hamburg, Germany.
Magnetom Impact Plus, Siemens Medical Solutions, Erlangen, Germany.
Digitizer ADC compact, AGFA, Leverkusen, Germany.
Scopix LR 5200, AGFA, Leverkusen, Germany.
Domitor, Pfizer GmbH, Karlsruhe, Germany.
Ketasel-5, Selectavet, Weyarn-Holzolling, Germany.
Antisedan, Pfizer GmbH, Karlsruhe, Germany.
EFILM workstation, version 1.9.4, Merge Healthcare, Milwaukee, Wis.
PROC GLM, SAS, version 9.3, SAS Institute Inc, Cary, NC.
PROC NESTED, SAS, version 9.3, SAS Institute Inc, Cary, NC.
References
1. Kuoni W, Augustiny N, Rübel A. Application of magnetic resonance imaging in reptile medicine. MAGMA 1993; 1:61–63.
2. Jacobson ER. Reptiles. Vet Clin North Am Small Anim Pract 1987; 17:1203–1225.
3. Redrobe SP. Ultrasound of exotic species. In: Mannion P, ed. Diagnostic ultrasound in small animal practice. Oxford, England: Blackwell, 2006;301–331.
4. Valente AL, Cuenca R, Parga ML, et al. Cervical and coelomic radiologic features of the loggerhead sea turtle, Caretta caretta. Can J Vet Res 2006; 70:285–290.
5. Isaza R, Ackermann N, Jacobson ER. Ultrasound imaging of the coelomic structures in the boa constrictor (Boa constrictor). Vet Radiol Ultrasound 1993; 34:445–450.
6. Sainsbury AW, Gili C. Ultrasonographic anatomy and scanning technique of the celomic organs of the Bosc monitor (Varanus exanthematicus). J Zoo Wildl Med 1991; 22:421–433.
7. Holland MF, Hernandez-Divers S, Frank PM. Ultrasonographic appearance of the coelomic cavity in healthy green iguanas. J Am Vet Med Assoc 2008; 233:590–596.
8. Bucy DS, Guzman DS, Zwingenberger AL. Ultrasonographic anatomy of bearded dragons (Pogona vitticeps). J Am Vet Med Assoc 2015; 246:868–876.
9. Penninck DG, Stewart JS, Paulmurphy J, et al. Ultrasonography of the California desert tortoise (Xerobates agassizi)—anatomy and application. Vet Radiol 1991; 32:112–116.
10. Casares M. Untersuchungen zum Fortpflanzungsgeschehen bei Riesenschildkroeten (Geochelone elephantopus und G. gigantea) und Landschildkroeten (Testudo graeca und T. hermanni) anhand von Ultraschalldiagnostik und Steroidanalysen im Kot. Zool Gart NF 1995; 65:50–76.
11. Schildger BJ, Tenhu H, Kramer M, et al. Ultrasonography in reptiles. Berl Munch Tierarztl Wochenschr 1996; 109:136–141.
12. Hernandez-Divers S, Hernandez-Divers S. Diagnostic imaging in reptiles. In Pract 2001; 23:370–391.
13. Silverman S, Janssen DL. Diagnostic imaging (special techniques and procedures). In: Mader DR, ed. Reptile medicine and surgery. Philadelphia: WB Saunders Co, 1996;258–264.
14. Stetter MD. Diagnostic imaging of reptiles. In: Bonagura JD, ed. Kirk's current veterinary therapy XIII small animal practice. 13th ed. Philadelphia: WB Saunders Co, 2000;1163–1168.
15. Wyneken J. Contemporary computed tomography and magnetic resonance imaging of reptile anatomy. In: Mader DR, Divers S, eds. Current therapy in reptile medicine and surgery. St Louis: Elsevier Health Sciences, 2013;93–106.
16. Mathes KA, Günther P, Kowaleski N, et al. Im Beckenbereich festsitzend Harnstein bei einer Maurischen Landschildkroete (Testudo graeca). Tieraerztl Prax Ausg K Kleintiere Heimtiere Tieraerztliche Prax Kleintiere 2009; 37:427–432.
17. Silverman S. Diagnostic imaging. In: Mader DR, ed. Reptile medicine and surgery. St Louis: Saunders Elsevier, 2006;471–489.
18. Rübel A, Kuoni W, Augustiny N. Emerging techniques. CT scan and MRI in reptile medicine. Semin Avian Exotic Pet Med 1994; 3:156–160.
19. Straub J, Jurina K. Magnet resonance imaging in chelonians. Semin Avian Exotic Pet Med 2001; 10:181–186.
20. Pees M, Ludewig E, Plenz B, et al. Imaging diagnosis—seminoma causing liver compression in a spur-thighed tortoise (Testudo graeca). Vet Radiol Ultrasound 2015; 56:E21–E24.
21. Raiti P, Haramati N. Magnetic resonance imaging and computerized tomography of a gravid leopard tortoise (Geochelone pardalis pardalis) with metabolic bone disease. J Zoo Wildl Med 1997; 28:189–197.
22. Pye GW, Jacobson ER, Newell SM, et al. Serratia marcescens infection in a gopher tortoise, Gopherus polyphemus, and use of magnetic resonance imaging in diagnosing systemic disease. Bull Assoc Reptil Amphib Vet 1999; 9:8–11.
23. Rasmussen AS, Lauridsen H, Laustsen C, et al. High-resolution ex vivo magnetic resonance angiography: a feasibility study on biological and medical tissues. BMC Physiol 2010; 10:1–8.
24. Lauridsen H, Hansen K, Wang T, et al. Inside out: modern imaging techniques to reveal animal anatomy. PLoS One 2011; 6:e17879.
25. Erlacher-Reid CD, Norton TM, Harms CA, et al. Intestinal and cloacal strictures in free-ranging and aquarium-maintained green sea turtles (Chelonia mydas). J Zoo Wildl Med 2013; 44:408–429.
26. Meyer J. Roentgenologische, computertomographische und kernspintomographische Untersuchungen zum Verdauungstrakt der Griechischen Landschildkroete (Testudo Hermanni). In: Veterinaermedizinische Universitaet Wien Universitaetsklinik fuer Roentgenologie. Vienna: Veterinaermedizinische Universitaet Wien, 1996;197.
27. Summa N. Anatomical atlas by magnetic resonance imaging (MRI) of the red-eared slider turtle. Trachemy scripta (Schoepff, 1792). Nantes, France: La Faculté de Veterinaire Médecine de Nantes, Oniris école nationale vétérinaire agroalomentaire et de l'alimentation Nantes–Atlantiques, 2010;206.
28. Effer N. Zur Schnittbildanatomie der griechischen Landschildkroete (Testudo hermanni boettgeri) in der Magnetresonanztomografie: Physiologische und pathologische Befunde sowie transhibernale Veraenderungen unter definierten Bedingungen. Gieβen, Germany: Justus-Liebig-Universitaet Gieβen, 2004;202.
29. Croft LA, Graham JP, Schaf SA, et al. Evaluation of magnetic resonance imaging for detection of internal tumors in green turtles with cutaneous fibropapillomatosis. J Am Vet Med Assoc 2004; 225:1428–1435.
30. Valente AL, Cuenca R, Zamora MA, et al. Sectional anatomic and magnetic resonance imaging features of coelomic structures of loggerhead sea turtles. Am J Vet Res 2006; 67:1347–1353.
31. Summa NM, Risi EE, Fusellier M, et al. Magnetic resonance imaging and cross-sectional anatomy of the coelomic cavity in a red-eared slider (Trachemys scripta elegans) and yellow-bellied sliders (Trachemys scripta scripta). J Herpetological Med Surg 2012; 22:107–116.
32. Arencibia A, Hidalgo MR, Vázquez JM, et al. Sectional anatomic and magnetic resonance imaging features of the head of juvenile loggerhead sea turtles (Caretta caretta). Am J Vet Res 2012; 73:1119–1127.
33. Mans C, Drees R, Sladky KK, et al. Effects of body position and extension of the neck and extremities on lung volume measured via computed tomography in red-eared slider turtles (Trachemys scripta elegans). J Am Vet Med Assoc 2013; 243:1190–1196.
34. Winer BJ. Design & analysis of single-factor experiments. In: Winer BJ, Brown DR, Michels KM, eds. Statistical principles in experimental design. New York: McGraw-Hill, 1971;149–260.
35. Ramsey PH. Power differences between pairwise multiple comparisons. J Am Stat Assoc 1978; 73:479–485.
36. Wilkinson R, Hernandez-Divers S, Lafortune M, et al. Diagnostic imaging techniques. In: McArthur S, Wilkinson R, Meyer J, eds. Medicine and surgery of tortoises and turtles. Oxford, England: Blackwell Publishing Ltd, 2004;187–238.
37. Blackband SJ, Stoskopf MK. In vivo nuclear magnetic resonance imaging and spectroscopy of aquatic organisms. Magn Reson Imaging 1990; 8:191–198.
38. Silverman S. Diagnostic imaging of exotic pets. Vet Clin North Am Small Anim Pract 1993; 23:1287–1299.
39. Holz P, Holz RM. Evaluation of ketamine, ketamine-xylazine and ketamine-midazolam anesthesia in red-eared sliders (Trachemys scripta elegans). J Zoo Wildl Med 1994; 25:531–537.
40. Greer LL, Jenne KJ, Diggs HE. Medetomidine-ketamine anesthesia in red-eared slider turtles (Trachemys scripta elegans). Contemp Top Lab Anim Sci 2001; 40:8–11.
41. Stoskopf MK. Clinical imaging in zoological medicine: a review. J Zoo Wildl Med 1989; 20:396–412.
42. Wilkinson R. Clinical pathology. In: McArthur S, Wilkinson R, Meyer J, eds. Medicine and surgery of tortoises and turtles. Oxford, England: Blackwell Publishing Ltd, 2004;141–186.
