General Abdomen

Bowel

Sonography is not routinely used as a primary tool for evaluation of the bowel. Nevertheless, there are many patients with nonspecific bowel-related complaints who are initially scanned with ultrasound, and in these patients attention to the intestinal structures can often identify the abnormality and direct the workup in the appropriate direction. Therefore a quick survey of the bowel is a useful undertaking in patients undergoing abdominal sonography. When necessary, sono­graphy can be performed following ingestion of water to improve evaluation of the stomach and proximal small bowel. Retrograde infusion of water can also be used to distend the colon (hydrocolonic sonography) and evaluate colonic lesions.

The normal bowel has five layers that can be seen sonographically. The inner hyperechoic layer arises from the interface reflection between the lumen and the surface of the mucosa. The second layer is hypoechoic and arises from the combined mucosa and the muscularis mucosa. The third layer is hyperechoic and arises from the submucosa. The fourth layer is hypoechoic and arises from the muscularis propria. The final outer layer is hyperechoic and arises from the interface reflection between the muscularis propria and the serosa or adventitia (plus peri-intestinal fat). These five layers are routinely seen on endoscopic ultrasound and are intermittently seen on transcutaneous scans ( Table 9-1 and Fig. 9-1 ). On many transabdominal scans when the lumen is collapsed, the mucosa, the muscularis mucosa, and the submucosa cannot be distinguished into separate layers. This produces a central echogenic region that is surrounded by the hypoechoic muscularis propria, resulting in the typical bull's-eye appearance ( Fig. 9-2 ). In other cases the mucosa and the muscularis mucosa combine to form a central hypoechoic region, followed by the hyperechoic submucosa and then the hypoechoic muscularis propria ( Fig. 9-3 ). The normal intestinal wall should be less than 5 mm in thickness.

One of the most common intestinal abnormalities seen on sonography is intestinal obstruction. Distended, fluid-filled, peristalsing loops are the hallmark of bowel obstruction ( Fig. 9-4 and Video 9-1 ). In many cases, the gas-filled loops located in the nondependent anterior abdomen will obscure these findings. This pitfall can be avoided by scanning in the lateral flanks where dependent fluid-filled loops are located. When a bowel obstruction is detected, a search for the cause should be performed. Hernias, intrinsic bowel wall lesions, abdominal masses, and abdominal fluid collections can all be detected with sonography.

Colon cancer is the most common intestinal malignancy visualized with sonography. Cancers of the colon typically produce focal, irregular thickening of the wall that is circumferential and may be either symmetric or asymmetric ( Fig. 9-5 ). Colon cancer can easily be detected by scanning the colon in short axis, progressing from the normal proximal segment, into the abnormal segment, and then into the normal distal segment ( e-Fig. 9-1 and Video 9-2 ). Gastric cancers are also visible sonographically as a focal thickening. They are easiest to detect in the antrum and body ( Fig. 9-6 ). Small bowel cancers are much less common and are difficult to visualize sonographically.

Lymphoma tends to cause wall thickening that is concentric and involves long segments of the bowel ( Fig. 9-7 ). Although focal masses occur, they are less common. A characteristic of lymphoma is marked thickening of the small bowel with maintenance of a normal or dilated lumen. This is referred to as aneurysmal dilatation ( e-Fig. 9-2 and Video 9-3 ).

Gastrointestinal stromal tumors (GISTs) are submucosal lesions that tend to grow in an exophytic manner. In some cases, particularly large masses, GISTs appear as masses that lack a clear organ of origin. Large lesions often have central necrosis and calcification ( Fig. 9-8A and B ). In some cases the continuity with the bowel can be established (see Fig. 9-8C to F ). Rarely, the submucosal origin of the lesion is visible sonographically ( e-Fig. 9-3 and Video 9-4 ).

Polypoid lesions of the bowel are generally not seen on sonography. The exceptions occur in patients who are thin or when the polyp is large or echogenic such as with lipomas ( Fig. 9-9 ) ( e-Fig. 9-4 and Videos 9-5A and 9-5B ). Determination of the etiology of polypoid lesions usually depends on a combination of imaging features and clinical factors. Ultimately, most require tissue sampling for definitive diagnosis.

A wide variety of disorders can cause bowel wall thickening ( Box 9-1 ). Inflammatory thickening of the intestines is generally diffuse and concentric ( Fig. 9-10 ). Commonly encountered causes include diverticulitis, Crohn's disease, ischemia, colitis caused by Clostridium difficile , and ulcerative colitis. Although the sonographic findings are typically nonspecific, the combination of sonographic findings and clinical parameters is often sufficient to suggest the diagnosis. Sonography has assumed an important role in the evaluation of patients with Crohn's disease in many parts of the world. A combination of gray-scale findings in the small bowel wall and Doppler evaluation of vascularity can monitor the disease effectively ( Fig. 9-11 ). Increased use of ultrasound can reduce the radiation exposure associated with computed tomography (CT) and the costs associated with magnetic resonance imaging (MRI). Evaluation of the peri-intestinal structures for extraluminal abscess is important, especially in patients with Crohn's disease and diverticulitis ( Fig. 9-12 ).

Another condition that can produce bowel thickening is an intussusception. A close inspection will reveal multiple alternating hyperechoic and hypoechoic layers due to the presence of three overlapping mucosal and muscular layers of the intussuscipiens (distal segment) and the intussusceptum (proximal segment) ( Fig. 9-13 and Video 9-6 ). Inability to detect blood flow in the intussusception increases the likelihood of necrosis and predicts the need for surgery. Detection of blood flow is reassuring. In adults, approximately 90% of the cases are associated with a lead mass of some sort (e.g., polyp, lipoma, stromal tumor, lymphoma, metastasis, cancer). Other causes in adults include Meckel's diverticulum and sprue. With the increased use of CT and ultrasound, transient, idiopathic adult intussusceptions are being diagnosed more often. In children, intussusceptions are usually idiopathic.

Bowel wall pneumatosis is rarely seen sonographically. Nevertheless, because it can indicate bowel ischemia, it is important to recognize when encountered. The hallmark of bowel wall pneumatosis on sonography is the presence of echogenic reflectors that indicate gas in the dependent wall of the bowel ( e-Fig. 9-5 and Video 9-7 ). As elsewhere, ring-down artifacts will sometimes be present and confirm that the reflections are coming from gas.

Appendicitis is a common cause of acute abdominal pain and is the most common condition requiring urgent abdominal surgery. CT is the primary imaging modality used to image patients with suspected appendicitis, with sensitivity and specificity of approximately 90%. Sonography has sensitivity and specificity of approximately 80%. Sonography is used as the primary test in children and pregnant women to avoid radiation exposure, and in young women because of the frequency of gynecologic causes of pain. To be successful, a high-resolution probe (usually a linear array or occasionally a curved array operating at 5 MHz or higher) should be used. Graded compression is important to get as close to the appendix as possible. Graded compression also pushes bowel gas out of the way and makes it possible to determine whether the appendix compresses. It is helpful to have the patient localize the region of pain. In many cases the patient can point to a specific area of pain and this usually corresponds closely to the site of the abnormal appendix. The primary criterion for the diagnosis of appendicitis is an appendiceal diameter greater than 6 mm ( Fig. 9-14A ). In some patients an intraluminal appendicolith can be detected (see Fig. 9-14B ). Other associated findings ( Box 9-2 ) are inflamed, echogenic periappendiceal fat ( Fig. 9-15A ); loculated periappendiceal fluid collections; and hyperemia on color Doppler (see Fig. 9-15B and Video 9-8 ). Localized appendicitis isolated to the tip of the appendix can occur, and therefore it is important to follow the appendix to its blunt tip whenever possible ( Fig. 9-16 ). In women, transvaginal scans can supplement transabdominal scans for visualization of deep pelvic appendicitis ( Fig. 9-17 and Video 9-9 ). In some cases differentiating the appendix from loops of ileum can be problematic. The appendix should be blind ending and noncompressible, and peristalsis should be absent or very minimal ( Videos 9-10A and 9-10B ). It is important to realize that the normal appendix can be seen only by experienced sonographers, and usually only in thin patients ( Fig. 9-18 ). It is not necessary to see a normal appendix to exclude appendicitis.

Peritoneum

The most common abnormality of the peritoneal cavity is ascites. In addition to transudates and exudates, other less common considerations include urine, blood, pus, cerebrospinal fluid (related to shunts), peritoneal dialysis, and chyle. Sonography is quite sensitive at detecting ascites as well as guiding aspiration of ascites. Small amounts of ascites are seen as anechoic collections most often in the pelvic cul-de-sac and in the right upper quadrant between the liver and the abdominal wall or in the hepatorenal fossa (Morison's pouch) ( Fig. 9-19A and B ). Larger amounts of ascites will distribute throughout the peritoneal cavity. Uncomplicated ascites lacks internal echoes and although it may displace adjacent structures, ascites does not cause significant distortion of adjacent structures (see Fig. 9-19C ). Ascites due to blood (hemoperitoneum) has internal echoes and may appear solid initially (see Fig. 9-19D and E ). Over time the echogenicity of a hemoperitoneum decreases. In the proper clinical setting, mobile echoes within ascitic fluid are consistent with blood (see Fig. 9-19D ). However, low-level echoes do not necessarily indicate blood. Ascites can loculate into collections that have multiple septations and compress and distort adjacent structures (see Fig. 9-19F ).

One evolving use of ultrasound is in patients with blunt abdominal trauma. The major role of ultrasound is to identify a hemoperitoneum as a secondary sign of injury to an intra-abdominal organ ( Fig. 9-20 ). In trained hands, sonography is very effective in identifying a hemoperitoneum. Unfortunately, sonography is less effective in identifying the injured organ. Contrast-enhanced CT is substantially more sensitive in detecting organ injury and is just as sensitive at detecting hemoperitoneum. Therefore the use of ultrasound in a trauma patient is controversial. In the unusual situation when CT is not readily available or when the patient is too unstable to leave the trauma room, sonography is a valuable tool that can replace the diagnostic peritoneal lavage as a means of detecting hemoperitoneum. It can also be used as a means of serially following patients who have a normal examination or have nonsurgical abnormalities detected at initial imaging.

Many primary tumors can metastasize to the peritoneum, but the most common are gynecologic, gastrointestinal, pancreatic, bronchogenic, and breast tumors. Peritoneal metastases are often very small and are not detectable by any imaging technique. When lesions reach 1 cm in size, they can more reliably be detected with sonography and CT. On sonography, they typically appear as hypoechoic solid or complex nodules immediately deep to the abdominal wall ( Fig. 9-21A to C ). They can be distinguished from the adjacent loops of bowel by noting that they are spherical and do not communicate with other loops of bowel. Because they are typically superficial, high-resolution linear arrays or curved arrays can be used and focused at a level just below the deep abdominal fascia. Ascites is often present in patients with peritoneal metastases and can highlight the lesions as soft-tissue nodules or sheets of soft tissue lining the peritoneal surfaces (see Fig. 9-21D to G ) ( Videos 9-11A , 9-11B , and 9-11C ). As with other neoplastic processes, they usually have detectable blood flow and may be hypervascular (see Fig. 9-21H and I ). An unusual type of peritoneal implant is pseudomyxoma peritonei. This consists of mucinous implants and gelatinous peritoneal fluid arising from mucinous tumors of the ovary, gastrointestinal tract (especially the appendix), or rarely other sites. The appearance of pseudomyxoma peritonei is variable, ranging from loculated, anechoic fluid collections that exert mass effect on adjacent structures to septated collections with or without internal echoes to more echogenic masses that move like a gel ( Fig. 9-22 , e-Fig. 9-6 and Video 9-12 ). Mesothelioma, a rare primary malignancy of the peritoneum, can closely mimic peritoneal carcinomatosis ( Fig. 9-23 ), as can tuberculous peritonitis. Splenosis is also a cause of solid peritoneal masses that can simulate peritoneal metastases. They are more homogeneous than most metastatic lesions but are otherwise indistinguishable ( Fig. 9-24 ). They have been described in more detail in Chapter 8 . Box 9-3 lists the common causes of peritoneal masses.

Omental metastases may appear as discrete focal lesions that are masslike or may be broad and flat ( Fig. 9-25A and B ). More extensive thickening of the omentum is often called omental caking. Soft-tissue replacement of the omentum is generally hypoechoic (see Fig. 9-25C and D ). Infiltration without complete soft-tissue replacement may appear hyperechoic and will often produce marked attenuation of sound (see Fig. 9-25E and F ). The platelike shape of the omentum, its location anterior to the bowel, its lack of peristalsis, and its lack of bowel wall morphology are all features that distinguish the omentum from the bowel ( Video 9-13 ). An uncommon cause of a peritoneal mass that can present as an acute abdomen is segmental infarction of the omentum. This is most common in men and most often occurs on the right side. Edematous infiltration of the omentum results in an echogenic mass that may produce marked sound attenuation ( Fig. 9-26 ). This abnormality can be easily overlooked if one is not familiar with its subtle appearance.

Although sonography is not typically used to evaluate a suspected pneumoperitoneum, the diagnosis can be made sonographically. Free air appears as bright reflectors, usually with dirty shadowing and/or ring-down artifacts, located along the nondependent aspect of the peritoneal cavity. The reflections are positioned immediately adjacent to the deep abdominal fascia. They will move to the nondependent portion of the abdomen when the patient changes position and can typically be seen between the liver and the abdominal wall when the patient is in the left lateral decubitus position ( Fig. 9-27 and Video 9-14 ).

Abdominal Wall

Because it is superficial, high-resolution probes can be used to evaluate the abdominal wall. In some situations it can be difficult to determine whether deep abdominal wall lesions are within the peritoneal cavity or within the abdominal wall. One useful maneuver is to scan in the longitudinal plane and ask the patient to take deep breaths. The respiratory motion of the intra-abdominal structures (bowel, omentum, and mesentery) will help to identify the deep abdominal fascia and confirm that abdominal wall lesions are superficial to this layer ( Video 9-15 ).

Masses of the abdominal wall can arise from a number of etiologies ( Box 9-4 ). Rectus sheath hematomas are one of the more common causes of a palpable or a painful abdominal wall mass. They are most often spontaneous secondary to anticoagulation. They may also result from direct trauma or forceful contraction of the rectus muscles, or may occur following surgery. The bleeding may involve the rectus muscle, the rectus sheath, or both. Above the arcuate line, the hematomas are unilateral. Below the arcuate line, they can cross the midline and involve both rectus muscles. They usually appear as complex collections with solid and liquefied areas ( Fig. 9-28 ). Because they are contained in the muscle or the rectus sheath, they usually are lenticular. When rectus sheath hematomas dissect inferior into the suprapubic region, they can enter the prevesical space, assume a more spherical shape, and exert significant mass effect on the bladder.

The most common abdominal wall neoplasm is metastatic disease. The two most common primary malignancies are lung and melanoma. Metastases can occur in the muscles, in the subcutaneous tissues, or rarely at the site of surgery or percutaneous needle tracts. Metastases usually appear as hypoechoic, solid masses or complex solid and cystic masses ( Fig. 9-29 ). Calcifications can occur but are uncommon.

Abdominal wall lymphoma and other lymphoproliferative masses can closely mimic metastatic disease. They can localize in the subcutaneous fat or the abdominal wall musculature. Cystic components are very uncommon, but despite being solid, portions of the tumor can appear anechoic and simulate cystic components ( Fig. 9-30 ). Lymphoma can be very vascular.

Desmoid tumors are rare fibromatous lesions that do not metastasize but can be locally invasive and can recur locally following surgery. They can be sporadic or associated with familial adenomatous polyposis or Gardner syndrome. They arise in the abdominal wall musculature or fascia, but unlike metastases or lymphoma they are not centered in the subcutaneous fat. They are solid, hypoechoic tumors with detectable vascularity ( Fig. 9-31 ). The margins are often infiltrative.

The most common extrauterine/extraovarian site for endometriosis is the abdominal wall, occurring most often in the scars following cesarean sections. Scar endometriomas may be entirely solid or have cystic elements. The margins can be well-defined, lobulated, or infiltrative. Blood flow may or may not be detected ( Fig. 9-32 ). It is important to realize that scar endometriomas do not have the characteristic diffuse, homogeneous, low-level echoes seen with pelvic endometriomas. Given a lack of characteristic sonographic features for abdominal wall endometriomas, a high index of suspicion should exist whenever a lower abdominal wall mass is encountered in a woman. Ultrasound-guided biopsies are usually diagnostic.

Abdominal hernias occur in areas where the abdominal wall is weak, either naturally weak (i.e., inguinal) or due to surgery (i.e., incisional). Up to 50% of men will develop a hernia during their lives. Many hernias can be diagnosed and treated without imaging of any kind. When imaging is necessary, sonography can provide the information necessary for management in the majority of individuals. Incisional hernias are seen as a fascial defect with the bowel ( Fig. 9-33 and Videos 9-16A and 9-16B ), fat ( Fig. 9-34 and Video 9-17 ), ascites, or a combination of structures ( Fig. 9-35 and Video 9-18 ) protruding through the defect. Hernias that contain only fat or fluid may be difficult to distinguish from lipomas or fluid collections. Compression using probe pressure and scanning when the patient performs a Valsalva maneuver can in most cases distinguish these possibilities by confirming that the mass communicates with intra-abdominal contents (see Videos 9-16A , 9-16B , 9-17 , and 9-18 ). Inguinal hernias originate at the internal inguinal ring, which is immediately adjacent to the inferior epigastric artery. The inferior epigastric artery can be identified in almost all patients as it travels deep to the rectus muscle. When in doubt, the inferior epigastric artery can be followed from its origin off the distal external iliac artery. Direct inguinal hernias arise inferior and medial to the inferior epigastric artery, whereas indirect inguinal hernias arise superior and lateral and travel anterior to the artery. Dynamic maneuvers such as Valsalva and compression with the transducer should be used not only to confirm the presence of a hernia but also to determine whether a hernia is incarcerated or reducible ( Fig. 9-36 , Video 9-19 ). Mesh plugs used to repair some hernias produce an inflammatory mass. This mass appears as a low-attenuation solid lesion on CT and is often 18-fluorodeoxyglucose avid on positron emission tomography scans, thus simulating metastatic disease. On sonography, mesh plugs appear as densely shadowing lesions ( e-Fig. 9-7 ).