principles and instruments of diagnostic ultrasound
Post on 29-Oct-2015
Embed Size (px)
This chapter is an April 2005 revision of a chapter writtenin April 1998. During that period, the basics of ultrasoundexamination have been surprisingly stable. Electronicshave shrunk so that now full function ultrasound duplexscanners can t in your pocket. The speed of personalcomputers has advanced so that the new scanners are justsoftware residing in personal computers equipped withan ultrasound receiver on the front end and a printer orDICOM adaptor on the back end.
In spite of shrinking electronics, all of the moderntriplex-Doppler, color ow, (three-dimensional real-time)four-dimensional scanners have converged on a singlestandard package. Most instruments are still 20 inches (50cm) wide, 30 inches (75cm) long, and 50 inches (125cm) tall, with a power cord that requires 120V (240V in Europe), 60Hz (50Hz in Europe), delivering amaximum of 15 A (7.5 A in Europe) or 1800W (volts multiplied by amps) with a printer included.This standardwas developed because doorways are 30 inches (760mm)wide and the eyes of ultrasound examiners are located 60 inches (150cm) off the ground, and power outlets are ubiquitous. This way, an examiner can roll the ultrasound instrument through a door while the exam-iners view over the top of the instrument is unobstructedand connect the instrument to power at the patientsbedside. Thus the systems are portable, something thatwill always be out of reach of the computed tomography(CT) and magnetic resonance imaging (MRI) alternativetechnologies. It is truly surprising that the manufacturers,over the period from 1979 (when commercial duplexscanners became popular) to 2005, while electronics haveshrunk to (1/2)18 or 1/250,000 size, have elected to keepthe instrument package unchanged until this year with theintroduction of notebook size systems. What has changedis that the complexity of the systems has exploded. Nowultrasound scanners are over 4000 times as complex as in 1979. But, that era has come to an end. The rst companies to introduce palm size scanners have revertedto mounting them on standard size carts, not because
small size is impossible, but because the customers appearto want ultrasound scanners that look large. Now, all themajor manufacturers offer palm size ultrasound scannersas an alternative to the large systems.
This chapter was written in the United States in a worldwhere inches and feet, gallons and quarts are the rule,and room temperature is 72F (22C) and body tempera-ture is 98.6F (37C). Science has advanced to use adecimal metric system. The decimal system, however, ismisaligned with the computer-friendly binary numbersystem. As an example, 0.1 is an irrational number inbinary computer math. In the future, I predict, the decimalsystem will be discarded in favor of a binary (or hexa-decimal) system that is more compatible with computers.In the meantime, the relationship between the decimaland binary systems will require constant explanation.
In this chapter, the MKS (meter kilogram second)system will be used when alternatives are not more con-venient.To facilitate understanding, all units will be listedevery time a number is given. To make things more dif-cult, the scientists have decided to honor their own bynaming often used groups of units after the great peoplein physics, like Hertz for frequency instead of cycles persecond, Pascals for pressure instead of Newton persquare meter, and Rayls for impedance instead of kilo-grams per square meter per second. In this chapter, alongwith the modern names, the more primitive units ofmeasure will be included in order to foster the ability ofthe reader to correctly perform important computations,which facilitate the understanding of future instrumentsand methods. The results of many computations are quiteastounding.
Medical imaging of the body requires the completion ofthree tasks: (1) locating volumes of tissue in the body(voxels) to be examined, (2) measuring one or more
3Principles and Instruments of DiagnosticUltrasound and Doppler UltrasoundKirk W. Beach, Marla Paun, and Jean F. Primozich
12 K.W. Beach et al.
feature(s) of the tissue within each volume, and (3) deliv-ering the data to the brain of the examiner/interpreter.The primary sensory channel used in medical imaging is the retina, a two-dimensional, analog/trichromatic(color), cine-capable (motion) data path. A secondarychannel used exclusively in ultrasound Doppler is dual audio channels (two ears). Although used in physi-cal examination, the tactile, taste, and olfactory senses are not currently used in medical imaging. A two-dimensional medical image presentation involves twotasks: (1) arranging areas on a screen (pixels) that corre-spond to tissue voxels and (2) displaying the measure-ment from the corresponding tissue voxel in each pixel.If more than one measurement type is made from eachvoxel (for example, echogenicity and tissue velocity) then both data types can be shown in a single pixel onthe screen by differentiating the echogenicity (shown in gray scale) from the velocity (shown in colors) by usingthe trivariable (red-green-blue) nature of the eye. If the process is done once to form one image, the image is called static; if the process is repeated rapidly (10 or more times per second) to show motion, then theprocess may be called real time. These steps arerequired in all sectional imaging modalities such as CT,MRI, positron emission tomography (PET) imaging,and ultrasound. Note that the voxels used in sectionalimaging are different from the voxels used in projectionalimaging like conventional X-ray and nuclear imaging.In sectional imaging, the voxels are nearly cubic, whereasin projectional imaging the voxels are long rods pene-trating the entire body and viewed from the end as pixels(dots).
In the following discussion, ultrasonic imaging consistsof three steps: (1) the voxels will be located, (2) meas-urements on each tissue voxel will be made, and (3) thetissue data from each voxel in the image plane will be dis-played as a two-dimensional array of pixels. This discus-sion will be limited to pulse echo ultrasound; continuouswave ultrasound will not be covered. There are severalalternatives to the popular methods that may becomeuseful in the future.
Locating Voxels with Pulse Echo Ultrasound
A transmitting ultrasound transducer is designed todirect a pulse of ultrasound along a needle-like beampattern from the transducer through the body tissues.Voxels located along that beam pattern provide echoesthat return to the receiving transducer along a needle-like beam pattern. Usually the transmitting transduceraperture and the receiving transducer aperture are at thesame location, so the transmit beam path and the receive
beam path are the same. The location of each voxelreecting ultrasound from along the beam can be deter-mined by the time taken for the echo to return. Data froma reector at a depth of 1cm returns to the transducer in13s; data from a depth of 3cm returns in 40s; data from15cm returns in 200s. Except for the details, this is acomplete description of pulse-echo ultrasound.To under-stand the physics and consequences on the image in moredetail, we need to separate four coordinate directions intissue: (1) distance from the ultrasound transducer(depth), (2) lateral direction in the image (ultrasoundbeam pattern width), (3) thickness direction in the image(ultrasound beam pattern thickness), and (4) time ofimage acquisition (frame interval/sweep) in the cardiaccycle. The physics of each of these is different. Thus, eachof these has an associated resolution. We will considereach of these in sequence.
Distance from the Ultrasound Transducer Depth
Pulse-echo ultrasound involves transmitting a short pulseof ultrasound into tissue and then receiving the echoesthat return from the tissues located at each depth alongthe ultrasound beam pattern.The echo time, the time aftertransmission for an echo to return from a voxel at a par-ticular depth, can be computed from the speed of ultra-sound in tissue. The speed of ultrasound in tissue is calledC. In most body tissues C is about 1500 ( 80) m/s =150,000cm/s = 1.5mm/s.
To compute the time [t] required for an echo to returnfrom a voxel at a particular depth [d], the round trip dis-tance of travel must be used [t = 2d/C]. For an echo toreturn from a depth of 3cm (30mm), the time requiredis [t = 2 30mm/1.5mm/s = 40s]. The result of thiscomputation for a series of depths is given in Table 31.
In a typical ultrasound machine, the time when echoesare returning is divided into 200 divisions, each divisionbeing 1s long. Each division takes data from a 0.75-mmvoxel along the ultrasound beam pattern to provide datafor a pixel along the beam line on the screen. Each beamline on the screen is thin and straight corresponding tothe thin straight ultrasound transmit beam pattern andthin straight ultrasound receive beam pattern. Most ultra-
Table 31. Ultrasound echo time for depth.
Depth Time Vessel
0.75mm 1s1.5mm 2s Finger artery
15mm 20s Supercial vein3cm 40s Carotid artery9cm 120s Renal artery
15cm 200 s
3. Principles and Instruments of Diagnostic Ultrasound and Doppler Ultrasound 13
sound systems assume that the speed of ultrasound isexactly 1.54mm/s; unfortunately the speed in some softtissues is 7% above or below this value.
Speed of Ultrasound in Tissue
The speed of ultrasound in tissue is determined by thecharacter of tissue. From basic physics, the speed can becomputed; it is determined by tissue density () and stiffness (K). C = . Ultrasound speed is the squareroot of the ratio of the tissue stiffness and the tissuedensity. Both stiffness and density are different in differ-ent tissues and both vary with temperature. Temperatureis not an issue (unless you examine a lizard or snake)because the normal mammali