Introduction

Trauma is a significant public health problem, rep-resenting the third leading cause of death in theUnited States. Trauma is also the leading cause ofmortality in Americans under the age of 40. Withthe widespread availability of multidetector-rowcomputed tomography (MDCT) in trauma centers,the traditional workup of trauma patients haschanged. Blunt chest injuries are now frequentlystudied with MDCT to evaluate the aorta, andworkup of the blunt trauma victim with abdomi-nal injury is evolving with MDCT. MDCT now al-lows not only the detection of injuries but providesnew information on the severity of injuries withimproved detection of vascular injury manifestedby “active extravasation.” Until recently, patientswith a history of penetrating trauma went directlyto the operating room for surgical therapy with-out preoperative imaging. Today, MDCT is oftenperformed in patients with penetrating trauma inorder to best identify vascular injuries prior tosurgical intervention.

This chapter reviews the technique of MDCTand discusses major findings in abdominal MDCTin the trauma patient. The varied manifestationsof bleeding are emphasized. Common mistakesand pitfalls in interpretation are described alongwith a step-by-step technique for interpretation.

MDCT Utilization in the Trauma Patient

Workup of the blunt trauma patient has evolvedwith the advent of MDCT. In the past, a large num-ber of patients with deceleration injuries had ab-dominal CT in the search for solid-organ or hollowviscus injury, but few had a chest CT. With the in-stallation of MDCT scanners in major trauma cen-ters, the frequency of chest CT in the trauma pa-

tient has increased dramatically. MDCT providesthe capability to perform high-definition multipla-nar reconstruction (MPR) based upon thin sec-tions reconstructed from MDCT raw data. ChestMDCT with MPR effectively produces CT an-giograms, probably equal in quality to angiogra-phy (Fig. 1).A trauma surgeon can be provided de-finitive information concerning aortic injuries al-most immediately with MDCT without the addi-tional contrast load and invasiveness of tradition-al angiography. At many institutions, MDCT of thechest has almost completely replaced angiographyin the initial workup of patients with possible aor-tic injuries. Angiography is often relegated to therole of a problem-solving tool. An MDCT of thechest not only provides diagnostic informationconcerning potential aortic injuries but also evalu-ates lungs, pleura, and bones. MPRs of the thorax

V.1MDCT of Abdominal Trauma

Robert A. Halvorsen

Fig. 1. Computed tomography angiogram (CTA) of motor vehicleaccident victim demonstrating two pseudoaneurysms of the aorta(arrows), the smaller located on the anterior surface at the bottomof the aortic arch and the second located on the posterior proximaldescending thoracic aorta

186 MDCT: A Practical Approach

facilitate detection of rib and spine fractures(Fig. 2a)

Alternative Strategies in MDCT Acquisitions

While the introduction of MDCT has dramaticallychanged the way many thoracic injuries are evalu-ated, it has had a lesser impact on evaluation of theabdomen and pelvis. Optimal use of MDCT belowthe diaphragm has not yet been established. Evalu-ation of the abdomen and pelvis with MDCT iscurrently performed differently in different insti-tutions. Some centers, such as Massachusetts Gen-eral Hospital, advocate the use of a “whole-body”CT in the trauma patient [1]. They utilize a contin-uous scanning technique through the areas to bescanned, such as cervical spine, chest, abdomen,and pelvis. Other centers, such as ours, continue toperform separate MDCT data acquisitions for eachtype of CT. For instance, a patient who has multipletypes of CT studies in our Emergency Departmentwill have them performed sequentially but not

continuously, facilitating the optimization of con-trast enhancement timing and radiation dose. Se-quential rather than continuous scanning makespossible the use of different types of reconstruc-tion algorithms for different anatomical segments.

Alternative Means of MDCT Interpretation

Another variable is the availability of freestandingimage processing workstations. Workstationslinked to CT scanners allow the interpreting radi-ologist to take raw data at the workstation andmake customized MPRs. Off-line reconstructionhas the added benefit of allowing the technologistto move on to the next patient without waiting forthe CT computers to perform the MPRs. Alterna-tively, technologists can produce routine MPRs us-ing standard imaging planes, such as sagittal orcoronal planes. Our technologists obtain routinecoronal and sagittal MPRs in chest trauma patientsand also reconstruct an oblique sagittal MPRalong the plane of the aortic arch (Fig. 1). When

Fig. 2a-c. Computed tomography (CT) of high-speed motor vehicle accident vic-tim. a Coronal CT with bone windows readily identifies rib fracture (arrow). b Softtissue coronal view demonstrates large amount of subcutaneous emphysema(white arrow) and active extravasation in upper abdomen medial to spleen (blackarrow). c Lung window in coronal projection demonstrates medial and small api-cal pneumothorax (black arrows)

a b

c

V.1 • MDCT of Abdominal Trauma 187

obtaining MPRs using the 16-detector-row MDCTscanner computer, time costs are of interest. AnMDCT of the chest requires a 20/second scan time.Initial reconstruction at 3 mm is performed, andaxial images are sent to our PACS system for inter-pretation. This initial reconstruction requires3 min 20 s. Then, to obtain MPRs, the raw data isreconstructed at thinner intervals and MPRs areconstructed, requiring 7 min 10 s. Therefore, in or-der to prepare data obtained from an MDCT of thechest for interpretation, scan time (data acquisi-tion) is only 20 s, but total reconstruction time is10 min 30 s.

Changes in Interpretation Strategieswith MDCT

Interpretation of MDCT in trauma patients requiresattention to detail. The use of a rigorous routine inthe interpretation of these studies significantly di-minishes missed traumatic lesions [2].We routinelyreview all trauma CTs with five settings:• Lung window• Soft tissue window• Liver window• Bone window• MPR: sagittal and coronal multiplanar recon-

structions

Interpretation: Routine Approach

Following is a detailed routine for interpretationof abdominal and pelvic MDCT in trauma pa-tients:

Lung WindowsIn our experience, the most frequently overlookedfinding in trauma CT is a pneumothorax. We use

lung windows to search for pneumothorax as wellas pneumoperitoneum (Figs. 2c, 3, and 4). Not on-ly the entire chest, but also the abdomen and pelvisare scanned from top to bottom using lung win-dows for the detection of free intraperitoneal air,intraperitoneal air adjacent to bowel loops, andretroperitoneal air.

Soft Tissue WindowsAfter scanning from top to bottom using lung win-dows, we switch to soft tissue windows and scrollfrom bottom back to top. This primary initial softtissue survey is performed to search for free in-traperitoneal fluid consistent with blood in a trau-ma patient (Fig. 5) Intraperitoneal fluid in a traumapatient is most likely due to either solid-organ orbowel injury and is a good indicator of injury sever-ity. Careful attention is paid to the presence or ab-sence of free fluid in the pelvis, where smallamounts of fluid are easily overlooked.After the ini-tial survey of the abdomen for blood, individual or-gans are scrutinized. When intraabdominal fluid isencountered on an MDCT of a trauma patient, ananalysis of fluid density is extremely helpful. Clottedblood next to or adjacent to a bleeding site is calleda sentinel clot [4] (Fig. 6). This clotted blood will beof higher density than the more serous blood fur-ther away from the site of bleeding. Identification ofthe sentinel clot is helpful in identifying the site ofbleeding.

Spleen SurveyWe evaluate the spleen twice. First, we look withinthe splenic parenchyma for areas of low or highdensity. Low density can represent splenic lacera-tion or fracture. Fracture of a solid organ is de-fined as a laceration that extends from one capsu-lar surface to the other. Splenic lacerations are usually identified because of the hematoma withinthe splenic parenchyma (Fig. 6). Whenever a

Fig. 3. Computed tomography (CT) of motor vehicle accident vic-tim. Demonstrates bilateral pneumothoraces

Fig. 4a, b. Motor vehicle accident victim with pneumoperi-toneum. Demonstrates subphrenic air (arrows)

a b

188 MDCT: A Practical Approach

hematoma is identified in the spleen or in any por-tion of an abdominal and pelvic CT, one should al-ways search for active extravasation. Active ex-travasation of contrast material signifies arterialbleeding. Recognition of arterial bleeding or activeextravasation has dramatically increased with theintroduction of MDCT with rapid administrationof high-concentration contrast media (Fig. 7). In1989, Sivit et al. [5] reported the first demonstra-tion of active intraabdominal arterial bleeding in apatient with splenic rupture from blunt trauma.Gavant et al. were among the first to describe theusefulness of detecting active bleeding in predict-ing the need for surgical intervention [3]. Jeffrey etal. [6] and later Federle et al. [7] further character-ized and clarified the importance of active ex-travasation in helical CT. Detection of active ex-travasation on CT implies arterial bleeding and isusually considered an indication for splenic arteri-ography with possible embolization as alternativeto surgery.

Traditionally, splenic injuries have been classi-fied using a CT-based scoring or grading system[8]. Such grading systems may be misleading, as aminor injury may go on to a devastating delayedbleed:• Grade 1: Subcapsular hematoma or laceration

<1 cm• Grade II: Larger subcapsular hematoma or lac-

eration 1–3 cm• Grade III: Capsular disruption or laceration

>3 cm• Grade IVA: Shattered spleen or active extrava-

sation into spleen or subcapsular hematoma orpseudo aneurysm or arteriovenous fistula(Fig. 7)

• Grade IVB: Active intraperitoneal bleeding(Fig. 8)The severity predicted by traditional CT scor-

ing systems for solid-organ injury using the Amer-ican Association for the Surgery of Trauma(AAST) scoring system is controversial, with anumber of authors finding them unhelpful [8–11],while others find them of use in patients with mas-sive splenic injury, as most patients with GradesIVA or IVB splenic injury will require catheter em-bolization or surgery [12].

Liver SurveyThe liver is the most frequently injured organ intrauma patients in general when both blunt andpenetrating trauma is considered, while in blunttrauma patients, the spleen is the most commonly

Fig. 5. Free intraperitoneal blood (arrow) in pelvis in blunt-trau-ma victim

Fig. 6. Sentinel clot. Motor vehicle accident victim with spleniclaceration. Perisplenic blood has higher density [61 Hounsfieldunits (HU)] compared with perihepatic blood (23 HU). Higher-den-sity blood adjacent to the spleen is the “sentinel clot,” helping toidentify the source of bleeding. Lower-density blood adjacent tothe liver is more serous in nature

Fig. 7. Motor vehicle accident victim with splenic injury demon-strating active intrasplenic arterial bleeding

V.1 • MDCT of Abdominal Trauma 189

injured organ. Survey of the liver is similar to thatof the spleen, with an initial review of the deep he-patic parenchyma in the search for laceration orhematoma. A second review of each slice contain-ing liver is performed to evaluate the margin of theliver in the search for subtle lacerations and peri-hepatic blood. Finally, one should evaluate the rightparacolic gutter for small amounts of fluid. Occa-sionally in a patient with extensive respiratory mo-tion, subtle hepatic injuries will not be detectable,but perihepatic blood, especially in the upper rightparacolic gutter, will point to the site of injury.

With hepatic injuries, as with any solid organinjury, it is important to look for signs of active ex-travasation manifested as high-density contrastequivalent to arterial structures on the same slice(Fig. 9). But besides arterial injury, venous injury isof extreme importance in hepatic trauma.With liv-er injuries, it is essential to look for signs of hepat-

ic vein damage. Traumatic avulsion of the hapaticvein occurs in approximately 13% of liver injuries,often as a result of avulsion of the right hepaticvein from the inferior vena cava. Such vein damageis suggested on CT when lacerations extendaround the inferior vena cava or into the porta he-patis (Fig. 10). With venous injuries, active ex-travasation is usually not detected. As the liverparenchyma itself compresses the laceration of thevein, no large hematomas are encountered. Howev-er, if the patient goes to the operating room andthe surgeon elevates the liver, the tamponading ef-fect of the liver parenchyma against the bleedingsite is removed, and patients frequently exsan-guinate on the operating room table. Therefore, ifthere is a detectable deep injury in the liver nearthe hepatic veins or the inferior vena cava (Fig. 11),the surgeon should be alerted to the finding priorto any operative intervention. With venous in-juries, control of the inferior vena cava is obtainedprior to elevating the liver in order to preventexsanguination.

Pancreatic and Duodenal InjuryDuodenal hematomas may be subtle, with onlymild thickening of the duodenal wall. Paraduode-nal fluids often have a triangular, pointed shapeand suggest a tear of the serosal surface of the duo-denum (Fig. 12).

Pancreatic lacerations are often associated withduodenal injuries but can occur without CT-de-tectable duodenal hematoma. Pancreatic lacera-tions are often difficult to diagnose on the immedi-ate trauma MDCT. Traumatic pancreatic injuriesrequire time to produce edema within the pan-creas. The initial CT may fail to show pancreatic in-jury unless a laceration within the pancreas is largeenough to be visualized or there is peripancreatic

Fig. 8. Motor vehicle accident victim with severely injured spleenwith active intraperitoneal bleeding (arrow) and evidence of ashattered spleen

Fig. 9a, b. Large subcapsular hematoma with active arterial bleeding from hepatic injury. a Linear area of active extravasation (arrow)lateral to liver within subcapsular hematoma on axial computed tomography (CT). b Celiac artery angiogram demonstrates active extrava-sation (arrow) in subcapsular hematoma mimicking vessel

a b

190 MDCT: A Practical Approach

bleeding. Always look for fluid density between thepancreas and the splenic vein (Fig. 13) Normally,only fat is found between the pancreas and thesplenic vein. If fluid is visible, then either traumaticpancreatitis or actual bleeding on the posteriorpancreatic surface is present. More obvious pancre-atic injuries will be detected as a linear lacerationextending through the tissue of the pancreas. Pan-creatic injuries often occur slightly to the right orleft of midline in locations where the pancreas issheered against the side of the vertebral body.Therefore, lacerations typically occur either at thejunction of the head and body of the pancreas tothe right of the spine or within the body just to theleft of midline (Fig. 14). The severity of a pancreat-ic injury is predominantly dependent upon the sta-tus of the main pancreatic duct. Bruises to the pan-creas can often be treated conservatively. However,

Fig. 10. Hepatic laceration with extension to inferior vena cava(arrow)

Fig. 11. Hepatic laceration extends along right hepatic vein andits branches (arrow)

Fig. 12a, b. Periduodenal hematoma (arrow) with triangular shape (a). Retroperitoneal hematoma (arrow) ( b)

a b

Fig. 13. Motor vehicle accident victim with pancreatic laceration(white arrow) through body of pancreas. This is a typical locationfor a deceleration injury due to shearing of the pancreas along sidethe vertebral body

V.1 • MDCT of Abdominal Trauma 191

a laceration or transection of the main pancreaticduct usually requires a surgical repair.

Patients with duodenal or pancreatic injuryshould be monitored carefully for significant pan-creatic injuries. Even if a patient has only mildswelling of the pancreas or inhomogeneity in blunttrauma, a follow-up CT is often warranted. Serumamylase may be used to detect change in amylaselevel suggesting traumatic pancreatitis, although theinitial amylase obtained in the Emergency Roommay be misleading. For instance, patients who havebeen subjected to head and neck injury may have anelevated amylase because of salivary gland injury.And the initial amylase in a pancreatic trauma pa-tient may be normal while it may rise later.

In patients with questionable pancreatic in-

juries, especially with an elevating amylase on seri-al lab tests, a magnetic resonance cholangiopan-creatography (MRCP) or occasionally an endo-scopic retrograde cholangiopancreatography (ER-CP) may be useful to better assess continuity of themain pancreatic duct (Fig. 14).

KidneysWith renal trauma, CT findings include laceration,fracture, and perirenal blood or urine. Kidneyanalysis in the trauma patient is different fromthat of the liver or spleen. While perihepatic orperisplenic fluid is usually due to blood, peri-nephric fluid may represent either urine or blood.Therefore, when renal trauma is suspected, it is es-sential to obtain delayed CT images (Fig. 15) Gen-

Fig. 14a, b. Motor vehicle accident victim with shearing injury of pancreatic body. a Laceration of posterior aspect of pancreas (white ar-row) with blood separating pancreas from splenic vein (black arrow). b Endoscopic retrograde cholangiopancreatography (ERCP) demon-strates large area of extravasation (black arrow) from pancreatic duct (white arrows)

Fig. 15a, b. Left renal injury: patient status post motor vehicleaccident. a Initial computed tomography (CT) study demonstratedfluid anterior to left kidney (arrow) and unusual fullness of renalpelvis. b Five-minute delayed image demonstrates urinoma (ar-rows). Fluid adjacent to the kidney and in the renal pelvis is ex-travasated urine, not blood

a b

a

b

192 MDCT: A Practical Approach

erally, such delayed images should be obtained af-ter a sufficient length of time for contrast to havebeen excreted into the renal collecting system, al-lowing for the detection of urinomas. Typically, a3-min delay is adequate. Our routine includes ini-tial evaluation of the abdominal CT prior to re-moving the patient from the CT table. If the pa-tient demonstrates any abnormality in the kidneyregion, delayed imaging is obtained. If the patientis known to have hematuria prior to CT, then de-layed images are protocoled in prior to initiationof the CT examination.

Retroperitoneal StructuresWhile adrenal injuries are often associated withrenal injuries, solitary adrenal hematomas can oc-cur without detectable renal injury. Adrenal in-juries usually manifest as simple adrenal masses.

It is important to image the inferior vena cavain the detection of shock. Shock is identified on CTas a flat or slit-like inferior vena cava on at leastthree slices in the infrahepatic inferior vena cava.We typically look at the inferior vena cava at thelevel of the left renal vein (Fig. 16). Please note thata flat or slit-like inferior vena cava seen on one ortwo sections only may be simply due to a rapid in-spiration of the patient, sucking blood out of theabdomen into the thorax if the patient gasps dur-ing CT examination. Therefore, one should see anarrowed inferior vena cava on three slices to in-crease specificity of this finding.

We recently reviewed our experience ofshocked or hypotensive patients studied with CTin our Emergency Department and identified thata small spleen is an additional finding of hypoten-sion (Fig. 17). In a series of patients who were hy-potensive either in the ambulance during trans-portation to or on arrival in the Emergency Room,we found that mean spleen volume in hypotensive

patients was 142 cc. Following fluid resuscitation,the spleen in the same patients was noted to in-crease to a mean volume of 227 cc.

Hollow Viscus InjuryIdentification of bowel injury in a blunt traumapatient is difficult. Bowel injury is often not de-tectable on clinical examination and can easily beoverlooked on a CT study. Findings suggestive ofbowel injury include free intraperitoneal air, freeintraperitoneal fluid, and wall thickening of thebowel. Unfortunately, extraluminal gas has beenreported to be detectable on CT, with a range of46–63% [13–15].When a loop of bowel that is fluidfilled and does not contain air is ruptured, therewill be no initial release of gas into the peri-toneum. Therefore, a CT obtained soon after abowel injury will often fail to detect extraluminalgas. Extraluminal contrast has been reported as ahelpful finding in abdominal trauma CT. However,in one study, extraluminal contrast was detected inonly 19% of cases [16]. In our experience, extralu-minal contrast is infrequently identified. As CTscans are obtained more rapidly following abdom-inal trauma, the incidence of detectable extralumi-nal contrast has declined. Patients are scanned soquickly after arrival in the Emergency Room thatadministered contrast, either given orally or vianasogastric tube, often has not had time to reachthe site of bowel injury. Reports from the radiolo-gy literature suggest an overall sensitivity for bow-el injury ranging between 88% and 93% [16]. How-ever, in the nonradiology literature there havebeen reports of significantly lower accuracy rates.In a 1998 study from a large trauma center in Texasinvolving 19,621 patients, CT missed hollow viscusinjuries in 43% of children and 21% of adults [17].

Bowel injury has been studied in an experi-mental model by a group of English surgeons [18].

Fig. 16. Motor vehicle accident victim with hepatic and renal in-juries and hypotension demonstrating slit-like inferior vena cava(arrow)

Fig. 17. Motor vehicle accident victim with bleeding into thighfrom femoral fracture demonstrates slit-like inferior vena cava(black arrow) and small spleen due to hypotension

V.1 • MDCT of Abdominal Trauma 193

Their model used the pig and studied decelerationinjuries. Their experiment consisted of anes-thetized pigs thrown by a mechanical device into asolid object. They found that bowel injury oc-curred in 100% of the pigs when the speed at im-pact was 100 mph (161 kph) or greater. In a similarfashion to that of humans, they found that small-bowel injuries were twice as frequent as colonic in-juries. Small-bowel injuries consisted of perfora-tion or mesenteric avulsions while colonic injurieswere usually serosal tears. The authors hypothe-sized that the increased frequency of small-bowelinjuries was due to the fact that the small bowel ismobile and the colon is relatively fixed in theretroperitoneum.

Detection of intraperitoneal fluid is critical inthe identification of bowel injury [19]. However,peritoneal fluid seen in the trauma patient can beeither from traumatic or nontraumatic origin.Traumatic causes of intraperitoneal fluid includeblood from a solid organ injury, blood from a bow-el injury, bowel contents, and blood from a mesen-teric injury but also can be due to bile from a rup-tured gallbladder or biliary tree or urine from aruptured bladder. Peritoneal fluid seen in the trau-ma patient can also arise from a combination ofmore than one injury. There are also nontraumaticcauses of intraperitoneal fluid, which can be quiteproblematic. The most difficult cause is encoun-tered in women of childbearing age who have asmall amount of what is termed “physiologic fluidin the pelvis.” In an ongoing study at our institu-tion, we reviewed 175 CTs of women of childbear-ing age who were referred for evaluation of blunttrauma. Of those patients who had no evidence ofinjury on CT and required no operative manage-ment of abdominal injury, approximately 50%were identified to have at least a small amount ofintraperitoneal fluid. In order to better character-ize this fluid, density, volume, and location wereanalyzed. Using CT reconstruction at a 5-mm in-terval, of the 175 patients, only one had fluid seenon more than three slices in the pelvis with no evi-dence of injury. Therefore, identification of a“trace” amount of fluid seen on less than threeslices seems likely to be an adequate predictor of anontraumatic, physiologic fluid collection.

Fluid location is also extremely helpful in iden-tifying the site of bleeding.As discussed above, theconcept of a sentinal clot is quite useful. Sinceblood clots adjacent to the site of bleeding, when ahigher density blood collection is encounteredwith a density that approaches that of adjacentmuscles, the site of bleeding is likely to be adjacentto this clotted blood. More serous blood is seenfurther away from the bleeding site.

With solid-organ injuries initial bleeding oc-curs adjacent to the injured organ then extendsdown the pericolic gutters and into the pelvis. On-

ly after readily accessible potential spaces arefilled, fluid will extend between the leaves of themesentery. With a large amount of intraperitonealblood from a solid organ injury “interloop” fluidwill be detected. However, if bleeding occurs due toa bowel injury, the initial bleeding initially occursinto the interloop space. Therefore, if a patient hasblood caught between the leaves of the mesenteryand does not have blood in the pericolic gutters orpelvis, then bleeding is likely to be from a bowelinjury.

Mesenteric or interloop fluid can be differenti-ated from the bowel by its shape [20], often mani-festing as V- or triangular-shaped fluid collectionsbetween the leaves of the mesentery that are easilydiscerned from the more rounded shape of fluidwithin bowel loops (Fig. 18). The etiology of the Vor triangular shape is simply that the mesenteryleaves converge at the root of the mesentery, andany fluid caught between the leaves tends to have apoint or apex of a triangle that points toward themesentery root (Figs. 18 and 19).

Another sign of possible bowel injury on ab-dominal and pelvic MDCT is a group of “matted-together” bowel loops (Fig. 20). This matted-to-gether appearance is due to blood extending be-tween loops of unopacified bowel. This is a rela-tively nonspecific appearance and can occasional-ly be seen in normal unopacified bowel but shouldbe considered a warning sign for possible bowelinjury.

Bowel-wall thickening is an important finding inbowel injury. Since most small- and large-bowelloops are not opacified with contrast on a CT ob-tained in a trauma patient, one must be able to iden-tify bowel-wall thickening without contrast. Oneuseful trick is to remember that bowel-wall thicken-ing is almost always circumferential in a trauma pa-tient. Therefore, look at the anterior portion of a

Fig. 18. Motor vehicle accident victim with extravasated urinefrom a bladder rupture: iodinated contrast caught between leavesof mesentery produces “V” shape (arrow) in upper pelvis

194 MDCT: A Practical Approach

bowel loop. While bowel contents may make thebowel wall look thickened posteriorly, often, there isenough air or fluid in the lumen to identify whetheror not there is anterior-wall thickening due to cir-cumferential injury. An additional trick is to callbowel-wall thickening only when it is seen in thesame bowel loop on two contiguous slices.

Identification of bowel injury is important. Inour experience at San Francisco General Hospital,46 patients with bowel injury had a delay in diag-nosis resulting in more than 6 h from the time ofinjury to operative intervention [13]. There was amortality rate of 4.3%, which corresponds to re-ports of mortality from bowel injury in the litera-ture, which is up to 5.9%. One of the two deathsthat occurred in our series was due to delayed di-agnosis of a single jejunal perforation. Delay in di-agnosis often occurred when the radiologist hadidentified an abnormality on the CT but the signif-icance of the CT findings was not appreciated. Forinstance, a radiology report described colonic wallthickening with pericolonic soft tissue stranding,but the impression failed to mention “possible” or“probably colonic injury.” In trauma patients withmultiple problems, it is quite useful to be specificin reporting possible bowel injury. While the find-ings on CT may be subtle, the consequence of a de-lay in diagnosis can be severe.

Conclusion

Since almost any portion of the abdomen andpelvis can be injured in a trauma patient, it is quiteuseful to use a routine interpretation techniquethat ensures that the radiologist reviews all appro-priate structures.

References1. Novelline RA, Rhea JT, Rao PM, Stuk JL (1999)

State-of-the-art. Helical CT in emergency radiology.Radiology 213(2):321–339

2. Halvorsen RA Jr, McCormick VD, Evans SJ (1994)Computed tomography of abdominal trauma; a stepby step approach. Emerg Radiol 1:283–291

3. Gavant ML, Schurr M, Flick PA et al (1997) Predict-ing clinical outcome of nonsurgical management ofblunt splenic injury: using CT to reveal abnormali-ties of splenic vasculature. AJR Am J Roentgenol168(1):207–212

4. Orwig D, Federle MP (1989) Localized clotted bloodas evidence of visceral trauma on CT: the sentinelclot sign. AJR Am J Roentgenol 153(4):747–749

5. Sivit CJ, Peclet MH, Taylor GA (1989) Life-threaten-ing intraperitoneal bleeding: demonstration withCT. Radiology 171:430

6. Jeffrey RB Jr, Cardoza JD, Olcott EW (1991) Detec-tion of active intraabdominal hemorrhage: value ofdynamic contrast-enhanced CT. AJR Am JRoentgenol 156:725–729

7. Federle MP, Courcoulas AP, Powell M et al (1998)Blunt splenic injury in adults: clinical and CT crite-ria for management, with emphasis on active ex-travasation. Radiology 206:137–142

8. Mirvis SE, Whitley NO, Gens DR (1989) Bluntsplenic trauma in adults: CT-based classificationand correlation with prognosis and treatment. Radi-ology 171:33–39

9. Umlas SL, Cronan JJ (1991) Splenic trauma: can CTgrading systems enable prediction of successfulnonsurgical treatment? Radiology 178:481–487

10. Kohn JS, Clark DE, Isler RJ et al (1994) Is computedtomographic grading of splenic injury useful in thenonsurgical management of blunt trauma? J Trau-ma Mar 36(3):385–389; discussion 390

11. Becker CD, Spring P, Glattli A, Schweizer W (1994)Blunt splenic trauma in adults: can CT findings beused to determine the need for surgery? AJR Am JRoentgenol 162:343–347

Fig. 19. Fluid caught between the leaves of the mesentery on thisdiagram demonstrates triangular shape (arrow)

Fig. 20. Motor vehicle accident victim with bowel injury demon-strating “matted-together loop” appearance (arrows)

V.1 • MDCT of Abdominal Trauma 195

12. Federle MP (2004) Splenic trauma. In: Federle, Jef-frey, Desser, Anne, Eraso (eds) Diagnostic Imaging:Abdomen. Amirsys, Salt Lake City, pp I:6:20–21

13. Harris HW, Morabito DJ, Mackersie RC et al (1999)Leukocytosis and free fluid are important indicatorsof isolated intestinal injury after blunt trauma. JTrauma 46:656–659

14. Mirvis SE, Gens DR, Shanmuganathan K (1992) Rup-ture of the bowel after blunt abdominal trauma: diag-nosis with CT. AJR Am J Roentgenol 159:1217–1221

15. Sherck J, Shatney C (1996) Significance of intraab-dominal extraluminal air detected by CT scan inblunt abdominal trauma. J Trauma 40:674–675

16. West OC (2000) Intraperitoneal abdominal injuries.In: West OC, Novelline RA, Wilson AJ (eds) Emer-

gency and trauma radiology. American RoentgenRay Society, Washington, pp 87–98

17. Allen GS, Moore FA, Cox CSJ et al (1998) Hollowviscus injury and blunt trauma. J Trauma 45:69–77

18. Cripps N, Cooper G (1997) Intestinal injury mecha-nisms after blunt abdominal impact. Ann R CollSurg Engl 79:115–120

19. Dowe MF, Shanmuganathan K, Mirvis SE et al(1997) CT findings of mesenteric injury after blunttrauma: implications for surgical intervention. AJRAm J Roentgenol 168:425–428

20. Halvorsen RA Jr, McKenney K (2002) Blunt traumato the gastrointestinal tract: CT findings with smallbowel and colonic injuries. Emerg Radiol 9(3):141–145