The clinical pathologist frequently encounters situations in which laboratory tests alone are not sufficient to provide a diagnosis. If this happens, certain diagnostic procedures may be suggested to provide additional information. These procedures are noted together with the laboratory tests that they complement or supplement. Nevertheless, it seems useful to summarize some basic information about these techniques and some data that, for various reasons, are not included elsewhere.
Diagnostic ultrasound
Ultrasound is based on the familiar principle of radar, differing primarily in the frequency of the sound waves. Very high-frequency (1-10 MHZ) sound emissions are directed toward an object, are reflected (echo production) by the target, and return to the detector, with a time delay proportional to the distance traveled. Differences in tissue or substance density result in a series of echoes produced by the surfaces of the various tissues or substances that lie in the path of the sound beam.
In A-mode (amplitude) readout, the echo signals are seen as spikes (similar to an electrocardiogram [ECG] tracing format) with the height of the spike corresponding to the intensity of the echo and the distance between spikes depending on the distance between the various interfaces (boundaries) of substances in the path of the sound beam. The A-mode technique is infrequently used today, but early in the development of ultrasound it was often used to examine the brain, since the skull prevented adequate B-mode (brightness-modulated, bistable) visualization.
In B-mode readout, the sonic generator (transducer) is moved in a line across an area while the echoes are depicted as tiny dots corresponding to the location (origin) of the echo. This produces a pattern of dots, which gives a visual image of the shape and degree of homogeneity of material in the path of the sound beam (the visual result is a tomographic slice or thin cross-section of the target, with a dot pattern form somewhat analogous to that of a nuclear medicine scan).
Gray-scale mode is a refinement of B-mode scan readout in which changes in amplitude (intensity) of the sonic beam produced by differential absorption through different substances in the path of the beam are converted to shades of gray in the dot pattern. This helps the observer recognize smaller changes in tissue density (somewhat analogous to an x-ray).
B-mode ultrasound (including gray-scale) is now the basic technique for most routine work. Limitations include problems with very dense materials that act as a barrier both to signal and to echo (e.g., bone or x-ray barium), and air, which is a poor transmitter of high-frequency sound (lungs, air in distended bowel or stomach, etc.).
In M-mode (motion) readout, the sonic generator and detector remain temporarily in one location, and each echo is depicted as a small dot relative to original echo location; this is similar to A mode, but it uses a single dot instead of a spike. However, a moving recorder shows changes in the echo pattern that occur if any structures in the sonic beam path move; changes in location of the echo dot are seen in the areas that move but not in the areas that are stationary. The result is a series of parallel lines, each line corresponding to the continuous record of one echo dot; stationary dots produce straight lines and moving dots become a wavy or ECG-like line. In fact, the technique and readout are somewhat analogous to those of the ECG, if each area of the heart were to produce its own ECG tracing and all were displayed together as a series of parallel tracings. The M-mode technique is used primarily in studies of the heart (echocardiography), particularly aortic and mitral valve function.
Real-time ultrasound is designed to provide a picture similar to B-mode ultrasound but that is obtained rapidly enough to capture motion changes. Theoretically, real-time means that the system is able to evaluate information as soon as it is received rather than storing or accumulating any data. This is analogous to a fluoroscope x-ray image compared to a conventional x-ray. M-mode ultrasound produces single-dimension outlines of a moving structure as it moves or changes shape, whereas real-time two-dimensional ultrasound produces an image very similar to that produced by a static B-mode scanner but much more rapidly (15-50 frames/second)—fast enough to provide the impression of motion when the various images obtained are viewed rapidly one after the other (either by direct viewing as they are obtained on a cathode ray tube (CRT) screen or as they are recorded and played back from magnetic tape or similar recording device). At present the equipment available to accomplish this exists in two forms: a linear array of crystals, the crystals being activated in sequence with electronic steering of the sound beam; and so-called small contact area (sector) scanners, having either an electronically phased crystal array or several crystals that are rotated mechanically. A triangular wedge-shaped image is obtained with the sector scanner and a somewhat more rectangular image with the linear-array scanner, both of which basically resemble images produced by a static B-mode scanner. On sector scanning, the apex of the triangle represents the ultrasound transducer (the sound wave generator and receiving crystal). The field of view (size of the triangle) of a typical real-time sector scanner is smaller than that of a standard static B-mode scanner (although the size differential is being decreased by new technology). Real-time image quality originally was inferior to that of static equipment, but this too has changed. Real-time equipment in general is less expensive, more compact, and more portable than static equipment; the ultrasound transducer is usually small and hand held and is generally designed to permit rapid changes in position to scan different areas rapidly using different planes of orientation. Many ultrasonographers now use real-time ultrasound as their primary ultrasound technique.
Uses of diagnostic ultrasound. With continuing improvements in equipment, capabilities of ultrasound are changing rapidly. A major advantage is that ultrasound is completely noninvasive; in addition, no radiation is administered, and no acute or chronic ill effects have yet been substantiated in either tissues or genetic apparatus. The following sections describe some of the major areas in which ultrasound may be helpful.
Differentiation of solid from cystic structures. This is helpful in the diagnosis of renal space-occupying lesions, nonfunctioning thyroid nodules, pancreatic pseudocyst, pelvic masses, and so on. When a structure is ultrasonically interpreted as a cyst, accuracy should be 90%-95%. Ultrasound is the best method for diagnosis of pancreatic pseudocyst.
Abscess detection. In the abdomen, reported accuracy (in a few small series) varies between 60% and 90%, with 80% probably a reasonable present-day expectation. Abscess within organs such as the liver may be seen and differentiated from a cyst or solid tumor. Obvious factors affecting accuracy are size and location of the abscess, as well as interference from air or barium in overlying bowel loops.
Differentiation of intrahepatic from extrahepatic biary obstruction. This is based on attempted visualization of common bile duct dilatation in extrahepatic obstruction. Current accuracy is probably about 80%-90%.
Ultrasound may be useful in demonstrating a dilated gallbladder when cholecystography is not possible or suggests nonfunction. It may also be helpful in diagnosis of cholecystitis. Reports indicate about 90%-95% accuracy in detection of gallbladder calculi or other significant abnormalities. Some medical centers advocate a protocol in which single-dose oral cholecystography is done first; if the gallbladder fails to visualize, ultrasonography is performed. Detection of stones would make double-dose oral cholecystography unnecessary. Some are now using ultrasound as the primary method of gallbladder examination.
Diagnosis of pancreatic carcinoma. Although islet cell tumors are too small to be seen, acinar carcinoma can be detected in approximately 75%-80% of instances. Pancreatic carcinoma cannot always be differentiated from pancreatitis. In the majority of institutions where computerized tomography (CT) is available, CT is preferred to ultrasound. CT generally provides better results in obese patients, and ultrasound usually provides better results in very thin patients.
Guidance of biopsy needles. Ultrasound is helpful in biopsies of organs such as the kidney.
Placental localization. Ultrasound is the procedure of choice for visualization of the fetus (fetal position, determination of fetal age by fetal measurements, detection of fetal anomalies, detection of fetal growth retardation), visualization of intrauterine or ectopic pregnancy, and diagnosis of hydatidiform mole. Ultrasound is the preferred method for direct visualization in obstetrics to avoid irradiation of mother or fetus.
Detection and delineation of abdominal aortic aneurysms. Ultrasound is the current method of choice for these aneurysms. Clot in the lumen, which causes problems for aortography, does not interfere with ultrasound. For dissecting abdominal aneurysms, however, ultrasound is much less reliable than aortography. Thoracic aneurysms are difficult to visualize by ultrasound with present techniques; esophageal transducers may help.
Detection of periaortic and retroperitoneal masses of enlarged lymph nodes. Ultrasound current accuracy is reported to be 80%-90%. However, CT has equal or better accuracy, and is preferred in many institutions because it visualizes the entire abdomen.
Ocular examinations. Although special equipment is needed, ultrasound has proved useful for detection of intraocular foreign bodies and tumors, as well as certain other conditions. This technique is especially helpful when opacity prevents adequate visual examination.
Cardiac diagnosis. Ultrasound using M-mode technique is the most sensitive and accurate method for detection of pericardial effusion, capable of detecting as little as 50 ml of fluid. A minor drawback is difficulty in finding loculated effusions. Mitral stenosis can be diagnosed accurately, and useful information can be obtained about other types of mitral dysfunction. Ultrasound can also provide information about aortic and tricuspid function, although not to the same degree as mitral valve studies. Entities such as hypertrophic subaortic stenosis and left atrial myxoma can frequently be identified. The thickness of the left ventricle can be estimated. Finally, vegetations of endocarditis may be detected on mitral, aortic, or tricuspid valves in more than one half of patients. Two-dimensional echocardiography is real-time ultrasound. It can perform most of the same functions as M-mode ultrasound, but in addition it provides more complete visualization of congenital heart defects and is able to demonstrate left ventricle heart wall motion or structural abnormalities in about 70%-80% of patients.
Doppler ultrasound is a special variant that can image blood flow. The Doppler effect is the change in ultrasound sound wave frequency produced when ultrasonic pulses are scattered by RBCs moving within a blood vessel. By moving the transducer along the path of a blood vessel, data can be obtained about the velocity of flow in areas over which the transducer moves. Most current Doppler equipment combines Doppler signals with B-mode ultrasonic imaging (“duplex scanning”). The B-mode component provides a picture of the vessel, whereas the Doppler component obtains flow data in that segment of the vessel. This combination is used to demonstrate areas of narrowing, obstruction, or blood flow turbulence in the vessel.
Computerized tomography (CT)
Originally known as computerized axial tomography (CAT), CT combines radiologic x-ray emission with nuclear medicine-type radiation detectors (rather than direct x-ray exposure of photographic film in the manner of ordinary radiology). Tissue density of the various components of the object or body part being scanned determines how much of the electron beam reaches the detector assembly, similar to conventional radiology. The original machines used a pencil-like x-ray beam that had to go back and forth over the scanning area, with each track being next to the previous one. Current equipment is of two basic types. Some manufacturers use a fan-shaped (triangular) beam with multiple gas-filled tube detectors on the opposite side of the object to be scanned (corresponding to the base of the x-ray beam triangle). The beam source and the multiple detector segment move at the same time and speed in a complete 360-degree circle around the object to be scanned. Other manufacturers use a single x-ray source emitting a fan-shaped beam that travels in a circle around the object to be scanned while outside of the x-ray source path is a complete circle of nonmoving detectors. In all cases a computer secures tissue density measurements from the detector as this is going on and eventually constructs a composite tissue density image similar in many aspects to those seen in ordinary x-rays. The image corresponds to a thin cross-section slice through the object (3-15 mm thick), in other words, a tissue cross-section slice viewed at a right angle (90 degrees) to the direction of the x-ray beam.
CT scan times necessary for each tissue slice vary with different manufacturers and with different models from the same manufacturer. The original CT units took more than 30 seconds per slice, second-generation CT units took about 20 seconds per slice, whereas current models can operate at less than 5 seconds per slice.
CT is currently the procedure of choice in detection of space-occupying lesions of the CNS. It is also very important (the procedure of choice for some) in detecting and delineating mass lesions of the abdomen (tumor, abscess, hemorrhage, etc.), mass lesions of organs (e.g., lung, adrenals or pancreas) and retroperitoneal adenopathy. It has also been advocated for differentiation of extrahepatic versus intrahepatic jaundice (using the criterion of a dilated common bile duct), but ultrasound is still more commonly used for this purpose due to lower cost, ease of performance, and scheduling considerations.
Nuclear medicine scanning
Nuclear medicine organ scans involve certain compounds that selectively localize in the organs of interest when administered to the patient. The compound is first made radioactive by tagging with a radioactive element. An exception is iodine used in thyroid diagnosis, which is already an element; in this case a radioactive isotope of iodine can be used. An isotope is a different form of the same element with the same chemical properties as the stable element form but physically unstable due to differences in the number of neutrons in the nucleus, this difference producing nuclear instability and leading to emission of radioactivity. After the radioactive compound is administered and sufficient uptake by the organ of interest is achieved, the organ is “scanned” with a radiation detector. This is usually a sodium iodide crystal. Radioactivity is transmuted into tiny flashes of light within the crystal. The location of the light flashes corresponds to the locations within the organ from which radioactivity is being emitted; the intensity of a light flash is proportional to the quantity of radiation detected. The detection device surveys (scans) the organ and produces an overall pattern of radioactivity (both the concentration and the distribution of activity), which it translates into a visual picture of light and dark areas.
Rectilinear scanners focus on one small area; the detector traverses the organ in a series of parallel lines to produce a complete (composite) picture. A “camera” device has a large-diameter crystal and remains stationary, with the field of view size dependent on the size of the crystal. The various organ scans are discussed in chapters that include biochemical function tests referable to the same organ.
The camera detectors are able to perform rapid-sequence imaging not possible on a rectilinear apparatus, and this can be used for “dynamic flow” studies. A bolus of radioactive material can be injected into the bloodstream and followed through major vessels and organs by data storage equipment or rapid (1- to 3-second) serial photographs. Although the image does not have a degree of resolution comparable to that of contrast medium angiography, major abnormalities in major blood vessels can be identified, and the uptake and early distribution of blood supply in specific tissues or organs can be visualized.
Data on radionuclide procedures are included in areas of laboratory test discussion when this seems appropriate.
Magnetic resonance imaging (MR or MRI)
Magnetic resonance (MR; originally called nuclear magnetic resonance) is the newest imaging process. This is based on the fact that nuclei of many chemical elements (notably those with an uneven number of protons or neutrons such as 1H or 31P) spin (“precess”) around a central axis. If a magnetic field is brought close by (using an electromagnet) the nuclei, still spinning, line up in the direction of the magnetic field. A new rate of spin (resonant frequency) will be proportional to the characteristics of the nucleus, the chemical environment, and the strength of the magnetic field. If the nuclei are then bombarded with an energy beam having the frequency of radio waves at a 90-degree angle to the electromagnetic field, the nuclei are pushed momentarily a little out of line. When the exciting radiofrequency energy is terminated, the nuclei return to their position in the magnetic field, giving up some energy. The energy may be transmitted to their immediate environment (called the “lattice,” the time required to give up the energy and return to position being called the “spin-lattice relaxation time,” or T1), or may be transmitted to adjacent nuclei of the same element, thus providing a realignment response of many nuclei (called “spin-spin relaxation time,” or T2). The absorption of radiofrequency energy can be detected by a spectrometer of special design. Besides differences in relaxation time, differences in proton density can also be detected and measured. MR proton density or relaxation time differs for different tissues and is affected by different disease processes and possibly by exogenous chemical manipulation. The instrumentation can produce computer-generated two-dimensional cross-section images of the nuclear changes that look like CT scans of tissue. Thus, MR can detect anatomical structural abnormality and changes in normal tissue and potentially can detect cellular dysfunction at the molecular level. Several manufacturers are producing MR instruments, which differ in the type and magnetic field strength of electromagnets used, the method of inducing disruptive energy into the magnetic field, and the method of detection and processing of results. Unlike CT, no radiation is given to the patient.