BME525_ch1

BME525_ch1 - )l ‘é 3.x 21.3:i.3.5:2 3 t l “i sit so...

Info iconThis preview shows pages 1–9. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: )l ‘é 3.x: 21% .3 :i: .3 .5: :2; 3 t l “i sit so. “‘33 ~ INTRODUCTION The study of medical imaging is concerned with the interaction of all forms of radiation with tissue and the development of appropriate technology to extracr clinically useful information from observations of this interaction. Such information is usually displayed in an image format. Medical images can be as simple as a projection or shadmv image—as first produced by Rontgen nearly 100 years ago and utilized today as a simple chest X-ray—or as complicated as a computer reconstructed image—as produced by computer» ized tomography (Cl‘) using X-rays or by magnetic resonance imaging (MRI) using intense magnetic fields. Although, strictly speaking, medical imaging began in 1895. with Rontgen's discoveries of X-rays and of the ability of X-rays to visualize bones and other structures within the living body [1], contemporary medical imaging began in the 197(1s with the advent of computerized tomography [2, 3]. Early, or what we call classical, medical imaging utilizes images that are a direct manifesta- tion of the interaction of some form of radiation with tissue. Three examples will illustrate. what we mean by classical imaging. First is the conventional X-ray procedure in which a beam of X-rays is directed through the patient onto a film. The developed film provides a shadow image of the patient which is a direct representation of the passage of X-rays through the body. Al- though such images are not quantitative, they do provide some measure of the attenuation of X-rays in tissue. Thus a section of soft tissue will appear darker than an equally thick section of bone, which attenuates more of the X-rays. It should be noted that even with current technological developments 0 ° .1... ..£.. Im-fiflm. ..£.. .mafiam. ..J=-..-¢amm_ ._5-.|mlfiflm_ mean-waning“- .a y. W mientional X-rat' “filming 5”” represent most medical facilities. I ‘ ‘ eitsm le of classrcal imaging, ‘ I fl ‘ ‘ l gdifgongdm-liriere a radioactive material is tmettcd tnto the patron In t , __ , and Its COLIISI: b IS ml.“ Ed 0H2! pdilfinl “1 a poor spatial resolution, its real it ' he time can iol cal function from t ’ y A K v “ It C m Fflhgzrlyoti‘be conventional nuclear medicine image Is a direct incasu v ' active isotope uscd‘ “11°”E's:lazigfnrgizlgineim;frigging, consider conventional medical Humming Hal-ga a pulse of ultrasonic energy is propaselctllmm [:9 fir-“3;; --;-'§¥E?andiii“he back'fiatiered Echo signal is recorded by the same trans) - angulating 0r moving the transducer (or by U§lfl3 3 lfaflfidnf“ diff!) llfihlf'fl? alltr" Sequential echo signals are recorded. and a cross~secttonal tm‘a‘gc‘o lle _§_ijec}.£§.-dispiayeddim” on a video monitor. UilraSOIJtld'll‘llflElfi-h \er "fill ig-mapping of ochoiniérleitics and are a dtrcct result of the Interactton ol t it. ultrasound pulse with tissue. ’ ‘ I In this test we will define modem or contemporary medical Imaging ' operationally as a two-pan process: (1) the collection of data concerning the. inteiraetionfiof some form of radiation with tissue. and (2) the transformation of these data into an image {or a set of images] using specific mathematical methods and computational tools. Note that our definitions for hoth classical ern ' i ' ant witbmour gencral definition of medical . u give, I of this cheater. Note aim that modern tmagtng can be represented as a generalization of classical imaging and tltat classical imaging is simply a specia1_:_.§:ase of modern imaging in which the ’ i (“:10hiiflireetly from the-}___intera"b”tion process. Whereas classical imaging ISW react and intuitive, modern imaging is indirect and, in many'cases, counter intuitive.._.__:§ince modern images are formed by processing, reformulating, or v ' Emmi! mi; .‘ {rpm the; asue/radiation interaction data base, the \ .a 15 is often refien'ie to as “reconstruction” and the image as a “recon- structed image.” The fim Elm“? capable 0f producingtrue reconstructed images was winged - i N gamma? in I??? at BM] in England. Hounsfield‘s -' _. ~ _, device was based in part on mathematical @Vfiifiped by A. M. Cormack [4] a decade earlier. For their cii'orts em and (bnnaclt were awarded the Nobel Prize in medicine in [979. “We Simplya CT imaging is based 0n the mathematical formalism that that. if an object is viewed frofl; a number of different angleS, then a isect‘gnaifmage ii Fa“ b8 Wadi-tiled (er “reconstructed”). 'I‘nus X«ray 8:; '5 as“? a mapping of X-ray attenuation m- The WWW“ REX”? 1972 represents the real beginning of ur concept. of imaging as merely rise of the radioimlope uptake. of the s the major imaging procedure at consider a conventional nuclear dv‘antage is that it provides a measure of 1145 BEGINNING. )t-mvs 5 Table 1.1 3-D image reconstruction algorithms ._.w. ............. "WW, ._.. _____ -_.—“WW g Parallel-Beam Mode i,__WWW_ ._. ..... .. é Fan— Noam Modc 3—D Projection Reconstruction 2—D and 3-D Pro: cction ; i fi—i) Projection | 2 § 5 Rea-"N'uc'mn l’arallci‘llcarn Mode locnnvrmion .............................. ........ “a é Cttnc-Bcttln Mode Algebraic Reconstruction Technique {ART} é [to rat ivc ' ------- --— ----------- —- , Momud Maxrmum Likelihood Reconstruction (\MLR) or i lispcctnlion Moximtzatiort (EM) Rcconstructmn q | Direct Fourier Reconstruction {QFR} ,. [vum'lcr = _ . ..................... .WW ............. .W .... .. [lei-"Unfiln'cmm ‘ Direct Fourier imaging (DH) in NMR 1 — . .1 taking a picture. It has also led to the development 01‘ 3-D imaging and is making quantitative imaging a reality. The application ol reconstructive tomography to conventional nuclear medicine imaging has led to the develop- ment of two new imaging modalities: single photon emission computed tomography iSi’iiC'l‘l and positron emission tomography {PET}, Similar applications to the laboraton technique. of nuclear magnetic resonance (NMR] has led to magnetic resonance imaging (MRI). The CT concept is currently being extended to 5-D magnetocnccphalography. electrical impedance tomography. and photon migration tomography, to name a few. Inherent to the development of these new imaging modalities has been the development of new reconstruction techniques. which are detailed in Table l—l. In this chapter we seek to provide a brief historical perspective for the various medical imaging modalities that are currently important. The various techniques are shown in Figs. 1—] and 1-2 where they are. characterized by the interrogation wavelengths. A parallel sequence. Will be followed in the suc- ceeding chapters which provide more detailed discussions of the various imaging modalities Although the various imagtng techniques will. 01' neces— sity. be treated separately, our goal is to provide a unified approach to the field of medical imaging. 1—1 THE BEGINNING WITH X-RAYS The history of medical imaging really began on November 8, 1895, when Wilhelm Konrad Riintgcn reported the discoch of what he called "‘a new lav-firml'flm .l II-KvIJv-erl'flm .l II-KvIJv-erl'flm .l II-KvIJv-erl'flm .l II-KvIJv-erl'flm .l II-Kvljv-u(:h.n.n F— . mom. X—Dand Visible Micrmnve UHF Photons Sol't X-ray Gamma ray 1 t W-I t f l tEV lll‘elflfl eV l~li1tt KEV I Merv CI’T PET SPECT Scintillation Camera II: ny (Film) X ray (CU F5511" N-ligTaith-“l-‘tn-n-o'safitfi P-tvfiii __.___ iliemwavglmaging bl R Imaging 1 t" ——g t——--o A rn rnrn urn I'll'l'l A 10‘ z A : Ionizing radiation for imaging. CPT. charged particle tomography; PET. positron ran-Maine SPECT, slngle photon emission computed tomography. m- represents the mnlonizinp‘ (notational-raging giyen tor comparison, " 1 CH: 106“: 1 MHz It!) (EH: .mv; —q~w—— M'agnctocnct’phalography Photon Migration Imaging Microwave Imaging HMR Imaging Ultrasound Figure 1»! Nonionlzlng radiation for imaging. required 2Q minutes. THE BEGINNINGng X-RAYS 7 kind of rays," :1 form of energy that. was more penetrating than anything previously described [1 I. He had serendipitoust discovered the phenomenon some months earlier and applied the term “X-rays“ to indicate its mysterious nature. By the time his initial report was published. he had spent nearly a full year in characterizing the new phenomenon. which became the focus of the remainder of his career. One of his earliest observations was that in using photographic dry plates sensitive to Xsrays. it was possible to exhibit a shadow of the bones of the hand. Rdntgen’s first (new famous) radiograph happened to be of the hand of his wife Bera. The earliest English translation of Rontgen’s 1895 paper appeared in the journal Nature on January 23, 1.896; this journal subsequently began running a special section entitled, "The Rdntgen Rays" as a designated forum to publish the vasr amount of experimental results on the phenomenon being submitted from scientists throughout the world. in an apparent fremy oi scientific enthusiasm, within 12 months of Réntgen’s initial paper, there appeared over 1000 publications related to the “Rontgen rays." Perhaps the most significant of these were subsequent papers by Rfinrgon rtth who has been described by scientific historians as one of the greatest caper-imam talists 01' his time. The impact of his discovery on medicine was obvious to Rfintgeu when he noted that the new ray could not only penetrate the human body, but that. difl’erent tissues were penetrated to different degrees, thus forming complex images on photographic glass plates or phosphor luminescent. Screens. Thus penetrability, as demonstrated on the X~ray photograph (which became known as the “Rtintgenogram”). was-a direct extensiou of Rontgen’s own observations. The reasons for the ditfercnees in tissue absorption in what has now become known as the “diagnostic range” of bray energies, specifically the photoelectric and Compton effects, were not understood until many..years later. Despite this lack of fundamental understanding of the mechanism underlying the interactions of X-rays with body tissues, the utility of the discovery and its farvreaching potential for medicine was clear to Rt’intgcn's contemporaries. The significance to the science of physics was appreciated as well. Rontgen received the first Nobel Prize in physics in 1901. Application to medicine quickly followed, and clinical use of Xstays for diagnosis soon became routine. The first clinical use of X-rays in the United States was on February 3, 1896. The patient was a young boy named Eddie McCarthy of HanOver, New Hampshire. who had fallen while ice skating two weeks earlier, injuring his wrist. His physician. Dr. Gilman Frost of Hanover, contacted his brother Edwin Frost who was a professor of astronomy at Dartmouth. Eddie was brought to the physics laboratory at Reed Hall, Dartmouth College, where Professor Edwin Frost used a battery-powered Crookcs‘ vacuum tube apparatus and a photographie glass plate to produce the first American Rt'intgenogram. revealing a colles‘ fracture. The exposure an...” -. .....a.wa .t .. ... -. _..enmnl a .4, -. _ Anal! . _..enM/n .t .. ... -._. 9 INTRODUCTION The first report of an experiment designed to demonstrate the ability of the Kenny-photograph to characterize tissue appeared in Nature within three months of Rfintgen's original paper. Cormack and Ingle. using two essentially identicalhuman finger bunes, demonstrated thar the x-ray opacity of bone was due to the concentration of calcium [5]. Front this a series of expertniculs designed to demonstrate the underlying nature of particular “shadows” ti.e., densities") and “blackenings” (i.e., “lucencles") on the images was devel- oped. The new technologies of radiography and fluoroscopy were of great. interest tothe American inventor Themas Alva Edison. Edison accepted a challenge from Randolph Hearst in 1896 to produce a radiograph 'of the living human brain. He- was not successful of course, due to the limitations or the. technique in discriminating the subtle differences in electron densmcs of the-different tissues. This limitation was partially ameliorated in subsequent years by the development of pneumocucephalography by W. Dandy in 1918, a technique in which intracranially injected air provides contrast to delineate the normally fluid-tilled compartments of the brain 16]. However: X-«ray imagingeéot‘ the actual brain was not accomplished until 19?2 with the introducmu of the CI‘ Scanner by its inventor Sir Godfrey Hounsficld of EMI, who was. in a very real sense. a successor of Edison. Significant advances in radiography occurred prior to World War I with the firs atiou of gastrointestinal contrast materials, such as ingested hismut . amine the stomach. Thereof course were technical improve— tiigents. Perhaps the most important of these was the development of equip» mom that could allow faster hugging times (on the order of seconds rather than minutes). that enabledahgtter;evaluation of complex moving tissues such as the lungs gflI'Ebral- angiography was serendipitously invented by the French phytiffitan‘SEgaz Moniz in 1927 while attempting to opacity the hrain in a fashion to the gallbladder by arterial injection of opaque contrast media. From the standpoint of physics and instrumentation, the first 3‘5 years of researchin radiology brought very few advances. There of course was a tremendous intprovement in the technical quality of the images. The underly« ing physics of X-ray production and absorption was clarified, pr0viding both a ‘theoretical framework and a practical basis for engineering improvement. Within medicine, a considerable body of empirical knowledge on medical image interpretation was also accumulated during this period. This knowl- edge was primarily derived from rigorous integration of the most fun- damental principles of differential X-ray absorption (e.g., the fact that air-containing structures appear less dense on Rontgenograms than do fat» containing structures or fluid~oontainiug structures) with classical anatomy and expanding knowledge on the anatomic pathology of the diseases detected and characterized on the images. The vast body of empirical knowledge led to the inception of a new medical specialty, which became known as Rontgeuology, or radiology. Spe» , t i it “3% ,i ' — ' _.m — .. . At ._ - ~—« 01H... mecca—u .fiilm—w‘v—va-fi-Pw‘ mama—.mmwu mum—.mme mama Wager—u w WCLEAR MEDICINE WITH RADIOACI'IVE ISOTOP‘ES 9 eialty societies and academic departments became established, and research in radiology received increasing governmental financial support. As a few]! 8 new scientitic discipline, radiological sciences. found a permanent home in universities throughout the world. Still imaging remained limited to two areas: (1) the “plain” radiograph. and (2} new applications of opaque contrast materials that served to outline the structure and physiologic func- tion of what would otherwise be radiographicallj.r indistinct tissue elements. It was not until the early 1950‘s. however, with the simultaneous development. of nuclear and ultrasonic imaging, that any truly new physical methods were developed. 1-2 NUCLEAR MEDICINE WITH RADIOACTIVE ISOTOPES In contrast to the science of radiography, the birmdate of nuclear isotope imaging (now known as nuclear medicine or: nuclear radiology) is less well defined. The very beginning can of crimson-be traced to the dimvery of natural radioactivity hy Antoine Henri .Becqurel in 1896, The of polouinm by Pierre Curie and Marja Sklodowska-Curie .in [893 mu fol- lowed. The significance of these contributions to pltysics-wassuch that these three scientists shared the third Nobel Prize in physics in; I903. The concept. of producing diagnostic images using radioactiye. materials is much more recent, as is the use of the term "nuclear medicine." which dates from the 19505 altd can be traced to the contributions of the nuclear chemist Paul Kohman who proposed the concept of the radionuclide as an atom. with a composition of its nucleus such that it had a measurable life Span (longer than lt)"‘“’ see). It may come as a surprise that the term “atomic medicine" was still in common use. in many clinical departments and in the literature of the field as recently as 1969. The initial primary application of radionuclidcs in medicine was not in diagnosis but in radiotltcrapeutic applications, including treatment of metastatic thyroid cancer hy radioactive iodine and the use of radium and radioactive isotopes ofcobalt and cesium [used primarily, but. not exclusively) for the treatment of malignant tumors. In fact the first reported medical use of a radioactive suhstance was by Eugene Bloch and the French physicians Henri [)anlos when they placed radium in contact with a tuberculous skin lesion. Although the concept of using radioactive tracers in physiologic research had heen introduced by George tie lnlevesy in 1923, the first report of the clinical use of the technique for imaging comes from G. E. Moore. reporting in thc journal Scream in 194?. Moore administered [-13] to patients in order to detect the presence of hrain tumors. This was not an imaging technique in the strict sense but rather a means of external, relatively noninvasive dctec» tion, Similarly in 1940 Benedict Cassen at the Universijalifornia, Los 10 INTRODUC‘I'I 0N Angeles, administered [~13] to compare the function of thyroid nodules in patients to normal thyroid tissue in volunteers, using crude instrumentation which he later refined in the early [9505 to produce the first rectilinear I scanner B]. These experiments by Cassen were arguably the true beginning of modern nuclear medicine. ‘ I I _ Teehnetium was discovered in 1937 by Perrier and Emilio Segre, and the metastable isotope 99m«’l‘c was first applied to clinical use in 196]. The history of this development has been well reviewed by Lindcman [8]. Tech- nctittm compounds soon became generally available for clinical diagnostic I use for a variety of applications (cg, the localization and characterization of tissues using technetium tracers) both in the form of pertechnitate salt. and as a radio label for a diverse group of phannaCcutical agents. A wide- variety of such Tc—labeled biologically active molecules known as “radiopharmaccuti— Gals” are now in everyday clinical use, Nuclear medicine techniques. deriving from the- physiologic localization of just such isotopically labeled, biologically active molecules to particular tissues, are intrinsically tissue—characterizing imaging techniques. ' (fine. of the most significant developments in imaging insrrumenration was the scintillation camera developed by Hal Anger in I952. [91. This scintilla- tion camera. known as an Anger camera, is still a workhorse for nuclear medicine, and constitutes a major component of nuclear medicine facilities even when expanded into mmputcrized or computed tomography methods such: ._ .sSiflQlC photon-emission tomography (Sl’ECI‘). Today SPECI‘ systems using two or moreefinger scintillation cameras are common [Uitlls for nuclear medicine imaging. Obviously, nuclear medicine visualizes physiological func- _‘_;_tj__ us or functional metabolisms in comparison to X»ray imaging in which 'tisfializatioa oi the structure is the main function. As will'be discussed in the computerized or computed tomography section, the introduction of Cl" not only stimulated the development of single photon lONOBFHDhY (SPECI'), but also positron emission tomography (PET). The first'itfilEEECT ol' the‘pre~C'l‘ period was developed by Kuhl and Edward in the early 19605 [10]. rl‘hcir system had four banks of detectors with linear and rotational scanning capability and provided the basic essential framework for the modern SPECT. It was a great success despite its inability to quantita- tively visualize the radionuclide distribution, apparently due to the lack of modern 3~D image reconstruction algorithms. During or even preceding Kuhl’s effort to visualize the radionuclide distribution using the scanning concept, Brownell and colleagues had long attempted to do the same. using positron-emitting radionuclides such as C“ and N” [11]. Parallel to Brownell’s effort. Rankowitz ct at. [12] had attempted to visualize the positron—emitting radionuclide distribution using a multidetcctor array in the form of a ring much the same way as today’s PET scanners. Although close to today’s SPECTS and PETS. all of those attempts from the so‘callcd prc-(“l' period were disappointing in the real usage of the methods and techniques. is /‘ . z ’23 z ..:M.:'.-.?£ssc.‘”m:‘. afi'fiiiix... mail! NUCLEAR uEDtctNE van-t WIN?“ ISOTOPES 1 1 Nevertheless, through those efforts much of the groundwork was established. When the CT concept was introduced in l9?2, it became clear that beth SPECI' and PET could be made fully quantitative devices using the newly introduced 3-D image reconsuuction algorithm [13]. The first wave of post-CT PETS were the PETI". a hcsaganal type detector array. developed by 'l‘er-Pogossian. Phelps. and Hofmann [14] at Washington University, and the circular ring type by Clio ct al. [[5] at UCLA and Budinger ct at. [to] at UC Berkeley. Those early PET systems. however, not only suffered poor resolution (resolution of the early systems was on the order of 2 cm l‘whnr [full width at half maximuml] but also sufiercd from a lack of sensitivity. Tbcy‘ captured only a fraction of the radiation emitted in 4n with relatively poor detection sensitivity using existing detectors. namely NaltTl). Great advances with the PET scanner were realized with the introduction of high 2 detectors such as EGO (bismuth germanat'e, Bifieyolyl in 197? [1?]. BGO not only improved resolution through use of small‘size detectors, it. improved sensitivity, especially the coincident detec- tion sensitivity which in effect improved more than an order of magnitude in comparison to Nalt'l‘l) of the same size. It soon became apparent that the detection geometry could be improved by using more rings or by extending the geometry to the cylindrical or even spherical multiple ring; systems. Today’s PET boasts resolution down to it - 6 mm t‘whm by utilizing small and narrow detecrors of 8- to lt‘i-layer rings. In the 19803 many more detection schemes were developed, such as the positology developed by ’I'anaka et a]. [18]. . There have also been great advances in SPEC’I' through better construc- tion of detector schemes and arrangements, such as three-head camera systems rather than the conventional two»head camera systems [19}. Resolu‘ tion of SPliCI" has been improved dramatically and is approaching 6 to 7 mm fwhm. The ultimate performance of SPECT. howasrer, still remains interior- to PET in resolution by almost a factor of two and sensitivity by as much as an order of magnitude, since the electronic collimation used in the PET is vastly Superior to the physical collimation employed in SI’IEZCF. The integration of Cl" methods into classical nuclear medicine imaging has dramatically expanded the range of diagnostic applications by introducing a greater degree of anatomic specificity and by providing for greater contrast resolution tban was previously available. Another noteworthy point is the radiopharmaccutical development. Classical. as well as contemporary nuclear medicine imaging. has benefited considerably from the development of nu- merous new radiopharmaceuticals. These chemicals have greatly expanded the range of organ systems that can be studied. In concluding this section. we should note that. whereas the goal of X‘ray imaging is to visualize structure, the objective of nuclear medicine imaging is to visualize physiological function or functional metabolism. Clearly such goals are complementary and not mutually exclusive. 1 2 INTRODUCTION 1-5 -X«RAY COMPUTED TOMOGHAPHY WITH THREE-DIMENSIONAL IMAGE RECONSTRUCTION The introduction of X—rtty computed tomography (or Cl‘) in t972 was perhaps the most revolutionary development in the field of medical imaging since the time of Rbntgen. For the first time the computer played a central role in the creation of thc images. The digital acquisition of data was in itself a revolutiOnary idea and brought with it a new concept. that 01 quantitativc imaging. Perhaps even more important was the new concept of image processing, the ability to alter and enhance the image after it was created, potentially extracting even more diagnostically relevant information. Anorltcr advantage of CT was that the method was tomographic and potentially three-dimensional, allowing the viewer to analyze isolated cross-sectional visual slices of the body. Technical advances with improved detectors. cont— puters, mathematics, and software, plus changes in mechanical components and system geometry, have produced images of muclt higher spatial and density resolation. Computed tomography allows ntuch liner diSL‘t‘imInntion of the intrinsic X—ray attenuation of the body tissues than was possible with plain radiograv phy. Typical equipmcnt allows discrimination of a range of tissue densities 1000-fold wider than was possible with even the best film-screen techniques. This new ability ted investigators to make use of the Quantitative (31‘ data in attempts to characterize. tissue and make. relatively noninvasive discrimina- tions between normal and pathologic tissues. The CI‘ density data was standardized by the introduction of the “EM! unit." which later became known as the “Honnsficld number," a calculated standardized index based on the mathematical estimate of the linear attenuation coefficient of the subjects-standardized with, reference to the measured attenuation coefficient of pure water. The I-lounslield scale was thus a fairly reproducible measure of the tissue’s ability to attenuate X-rays. The introduction of wmputed tomography engendered an enthusiastic response of basic and clinical investigators to apply the new technique for the narrates-of [Immechnracterizafion baSed primarily on the Hounstield number. loaaddition there were numerous published articles describing the cltaractcriw nation of tissues by measuring. the Hounslicld number of the tissue(sl in question foll0wing injection of intravenous iodinated contrast materials. Within three years of Hounsfield’s original report, over 1000 publications appeared in the English language imaging literature. The number of books and articles had more than qundmpled by 1.980, in a burst of enthusiastic research reminiseent of that which followed Riintgcn’s original report. Den Spite high hopes and expectations for the new technology, reliable discrimi- between normal and pathologic tissues by virtue of their Hotmsficld numbers was largely unsuccessful, except for a few and limited examples. interestingly, this very observation prompted a new line of experimenta- tion in which the energy of the x-ray beam was changed during the course of :g MAGHEHC RESONANCE imlNO mo WW 1 3 the srudy and a new image. derived from the mathematical subtraction or other manipulation of the two component images (obtained at different X«ray energies],‘was constmcted. This method became known as dualwenergy CT imaging; it has generated considerable interest with regard to CT-based tissue characterization, primarily with regard to quantification of high atomic number tissue elements, such as calcium in trabccular bone and pathologic tissue iron stores {20, 21]. I Much of the physics of contrast by the diti’erent energies of X-ray and tissue atomic composition has been studied by (Sho et at. [22] and also by McCullough [23]. Their studies have introduced in addition to the mean value of attenuation, a new index, thc “effective at0mic number” 2,, or :2. As mentioned earlier, another important area of research spurred by the developntent of X—ray CT was the image reconstruction algorithm, now known as 2H) and 3-D image reconstruction [see [13, 24, 25]; and Table 1-1) which will be discussed in detail in later chapters. For the past two decades 3-D image reconstruction mathematics has been the backbone of modern medical imaging and will likely remain so. 1-4 MAGNETIC RESONANCE IMAGING ANO TOMOGRAPHY It is rather a coincidence that in 'IWZ. the birthday of X~ray CT, a crude mode of NMR (nuclear magnetic resonance) imaging began to appear, or at least some preliminary form of NMR imaging commonly known now as MR1 (magnetic resonance imaging) had been suggested in the medical imaging community. At the time. however, it was premature and was greatly oven shadowed by the overwhelming Xeray Cl" “fever,” First. two papers appeared at nearly the same time: one by Paul C. Lautcrbur in 19?3 [26] and the other by Raymond Damadian in 1971 [2?]. At that time the phenomenon of nuclear magnetic resonance was not ncw. It had been discovered independently by Felix Bloch and Edward Purcell (for which the pair shared the 1952 Nobel Prize for physics) and extended by Richard R. Ernst whose work, using the Fourier transform of the raw NMR signal, was introduced in 1966 for production of N'MR spectra. Subsequently. largely inspired by P. C. Lauterbur’s work. a two-dimensional Fourier transform method using field gradtent pulses and phase encoding was introduced by Kumar, Welti, and Ernst in 19?5 [28], forming the basis of modern MRI. For this contribution Richard Ernst received the Nobel Prize in chemistry in 1991. Soon after, MR imaging front point mapping to the projection reconstruction type has gauged front many different research groups all over the world [29, 30, 31, The inherent advantages of MRI as an imaging tool are many. Chief among these are unprecedented contrasts between the various organs and tumors essential for image quality and the three-dimensio‘taturc of the- 14 INTRODUCTION method. Although there are a large number of different modalities within MR imaging—for example, flow imaging, T1 and T2 Weighted imaging, and spectroscopic imaging-#lhe single most important advantage‘of MR imaging is the contrast provided by the T, and Tl"2 relaxation mechanisms. As_we will elaborate later, MR imaging has now expanded into many different imaging modes—namely simple TI and T2 weighted imaging. chemical shift imaging for tissue discrimination such as normal tissue to fatty tissue. flow-related imaging including angiography or built- flow, and susceptibility cilihalnced dynamic functional imaging. Spectroscopic NMR and imaging ave a so )eeri developed extensively, especially in the fields. of proton and phosphorfia spectroscopies with both localization and imaging. A great future for M II also lies in the field of spectroscopic imaging, wherc’it might eventually provide tissue characterization with chemical spccrtieity. today. although not in common clinical use, many high-Field MRI scanners. above 1.5 tesla are capable of providing a limited usage of the systems for in-vivo spcciroscopic investigation of the human. body for medical diagnosis. I _ i ‘ In the beginning a somewhat. cumbersome feature of MR imaging was-its slow speed. Today the speed of imaging has been greatly improved. the fastest imaging times, which had been in the range of tens of minutes, are now reduced to as short as 50 nisec. Since MRI is capable of imaging magnetic susceptibility in relation to microvasciilarity, Ml} imaging rs now moving from static imaging to dynamic imaging such as microvaseular diffu- <sion and perfusion imaging. When this development is achieved, MRI will join in the competition with PET and SPECT in the race among dynamic physiologicaljunciional studies. 1~5 ULTRASOUND AND ACOUSTIC IMAGING The history-iots-imaging and tissue characterization by ultrasound has been agpquately reviewed elsewhere ([33, 34]). Here we only outlinesome-of the :iiio'dalities’ important features. Modern clinical ultrasound, including all '3'.§syStems currently in commercial use, is based on construction of 'dl'lxlfl‘liJg-C the-battltgcauered echo strength. These systems havc'their origin in World War l-Emavy sonar technology, from which the field ultimately derives. Such methods are limited in that theyr utilize only a fracticn of the acoustical information available in the echo wave form. In particular they exclude signal features which standard ultrasonic transducers are capable of detecting. Most technical advances in the field have served merely to improve upon the spatial and intensity discriminations of the equipment by producing improved transducers and imprevcments in post‘acquisition image‘cnhancement cir- cuitry. The additional signal data available has unfortunately hecu of only I an Gl'm1..fi.‘lfi -i'. I .I "'l fern—Jaw.” -i'. I .I "'l fern—Jaw.” Yak. . '>.: s \ : ,‘ casein“- .- "'l fern—Jaw.” ULTRASOUND mo ACOUSTIC motive 15 academic interest up to the present time. These “unused features" of the ultrasound signal, however. contain potentially useful information that is currently being discarded in routine clinical application. The interaction of sound with tissue structures is complex and is not fully understood; the science of ultrasound tissue characterization is derived from the belief that it may be possible to identify and utilize. particular features of the interaction of ultrasound energy that are characterisnc of certain lisancs. These features could then be utilized to make specilw tissue diagnoses noninvasively. Yet, despite years of disciplined experimentation and consid- erable advances in haste knowledge of the underlying physics of the p'rnpaga- tion of ultrasound in tissue, as well as in the interactions of high frequency sound waves with different types of tissues, no technique has evolved that will allow the desired end point of determination of specific "ultrasonic signa- tures“ by which specific tissues can be characterized. Scattering of the ultrasound beam: by inhombgeneities in the. tissue has received attention as a potential tissue—discriminating. property. Recently a number of investigators have demonstrated that scatter can be correlated with structural periodicity in the tissue sample, and they have found differ~ cnces in benign and malignant discascs of the liver in tissue samples that correlated with struttura] features evident in microscopy [3.5. 36, 37].- Other investigators have used back'scattered echoes to distinguish between”'nt)rrrial and diseased myocardimri [38, 39]. ' The frequency dependence of attenuation (due to changes in the scatter- ing process) was also studied, and a dilfcrence in myocardial infarction was found. Ultrasonnd attenuation in normal and infatcted myocadium has been studied for several years, and many differences between normal and is- chemic/infarcled tissues have been found.- For example. if there is an initial decrease in the attenuation in the early stages of the infarei, it was felt to be related to tissue edema. As Scar tissue forms, attenuation increases, probably as a reflection of the increased collagen matrix of the necrotic tissues. While data on ultrasound attenuation remain primarily empirical, without a comprehensive unifying theory, a large number of observations have re» vealed significant trends: (1) Attenuation increases with increasing collagen content and decreases with increasing water content. and (2) with the current diagnostic range of ultrasound frequencies, the frequency dependence of sound attenuation in tissue is approximately linear in soft tisSues. Conse- quently, tissues can be characterized, at least to Some extent. by application of sound wave velocity and attenuation measurements. Measurements of this sort are still being reported. The role of density and elasticity fluctuations in acoustical scattering was first described by Lord John Rayleigh ever 100 years ago with reference to the propagation of sound in air. This “classical” approach has value with regard to the propagation of sound in tissue as well. and to some degree it. can be described by classical acoustics. Assuming classical acoustical "'|'7€l"nr_1..fi.'lfi 'I'- I i' 'v'iflrl'mnmlt? 'I'- I i' "xterm—now.“ 'I'- I i' "xterm—now.“ 'I'- I i' "xterm—now.“ 'I'- I i' " 1 5 IN TROWCHON behavior, it is possible to calculate the effective impedance of sound of lltflrictus tissues. This concept was first introduced to describe tissue ultra- spund scattering in quantitative terms [40, 41] and recently it has provided a theoretical framework for the approach to ultrasonic tissue characterization by acoustical measurement [42]. Within the laSt few years there have been increased etforts to measure the “backseattering coefficient" of various tissue samples and to relate this parameter to the pathological state of the tissues. One method [43] utilizes the angle of ditfraetion—the Bragg scattering method, Which measures the variations of a particular frequency component of the backseatlering coeffi- cient of a small region of interest (interrogated from different directions). In theory-this index should reflect components of the tissue structure; despite its inherent merits and a great deal of interest iii the technique. its clinical application has been limited to date. v-..The. application of particular signal‘proceSsing prl’OHCheS to extract. tis» sue-specific information has been pursued by several investigators, all of whom have proposed using a post-aequisition homomorphie filtering algo- rithm to provide a measure of the frequency-dependent attenuation by estimation of changes in the shape of the ultrasound pulse. The method is complex computationally intensive, presenting a practical limitation in implementation [44]. However, it would allow accurate measurements of tissue sound impedance and attenuation ;;in a single measurcmcnt with a single index that could be compared between tissue types. A similar method involving an “adaptive filtering technique”:has been proposed and used [45]. _ Ferrariand Jones et at. in 1932 proposed an “FM” ultrasound system [46]. Thissystejit incomerated phasc information (the "FM" signal) along with reheated echo tutle (the “AM” signal) in thc generation of a single image that combined information on both the echo amplitude and its inherent phase information. By incorporating this phase information into the acoustical-.bnagemore Specific tissue characterization might be possible. This technique; was-unfortunately fraught with technical difficulties and has had ' i'DOCIl’ clinical acceptance. Very recently prototype equipment has been produced that employs very high frequency ultrasound to produce high-resolution images on a micro» scopic scale,- thereby creating a new science called “ultrasound microscopy,” The new technique requires extensive instrumentation, but it has shown promise in i‘inoninvasive" microscopy (i.e., untbted and largely unprepared tissue samples can be studied) as well as serving to extend the knowledge of how very high frequency ultrasound (in the gigahertz range) interacts with tissue [47, 48]. in limited clinical applications the images have been favorable compared to conventional fixed and stained tissue specimens in a variety of skin lesions. and their microscopic information content has proved to be of diagnostic quality similar to conventional hematoxalin/eosin stained tissue specimens [49]. REFERENCES 17 1-6 NEUROMAGNEI'IC lMAGING—AN EMERGING IMAGING MODALITY The recent development of SQUID (superconducting quantum interference device), with which one can mcasure magnetic field variations as small as 10"" tesla, will enable neuroscience researchers to observe accurately the site of neuronal activity in the brain and eventually to reconstruct the activity in three-dimensional form. SOUlD‘s fast temporal resolution and better localization capability. compared with methods such as EEG. makes it a unique method for measuring neuronal activity. Thus magnetoencephalogra- phy iMEG) is a potentially powerful tool for studying the spatiotemporal distribution of neuronal activity. More recent developments in neuromagnetic imaging ineludc the tomor- graphie imaging of the neuronal activity using 3-D reconstruction algorithms. Singh and Doria. in 1984 [50], were the first to develop and demonstrate the feasibility of imaging neuronal activity using CT-lilte-image reconstructiOn algorithms. Subsequently neuroniagnetic localization has been performed in conjunction With MRI. Although neuromagnctic research is a relatively new area, it has unprecedented pctential for examining and observing the various neuronal activities in the brain, thereby enabling one to Study brain functions with a time response much faster than any other imaging modalities. Cur- rently, with the advent of a 37-channel SQUID netuqmagnetometer which is commercially available and suitable pressessing techniques, it has hecorne possible to locate or image the center of the 3-D distribution of the'heuronal activity evoked by external stimulation such as vision and hearing REFERENCES I. W. K. Roentgen (trend. by A. Stanton). On a new kind of rays. Nature 53:34—36 (23 January 1896]. 2. G. N. ilounslield. A method of and apparatus for examination of a body by {jiggigiflfl such as X-ra)’ or gamma radiation. British Patent No. 1283915. London G. N. Hounslield. Potential uses of more accurate CT absorption values by filtering. AJR 131:103 (INS). - 4. A. hl. Cormack. Representation of a function by its line integrals. with some radiological applications. J. Appl. Physics 341722—32? (1963). 5. .l. D. Cormack and H. Inglc. The Roentgen rays. Nature 53:43? (12 March 1396). 6, w. E‘. Dandy. Ventriculography following the injection of air into the cerebral ventricles. Am. J. Surger 68:5 (NIB). '2'. B. Casscn, L. Curtis, C. Reed. and R. Libby. Instrumentation lot '“I used in medical studies. Nttcfemtt'cs 9:46 (I951). bl 'um -\' menu-mm rum.- -'.-t.uw.-.rm t'um -\' menu-sum t'um -\' t-rtuu-xum t'um -\' menu-mm t'um -\' t-t-tuu-mm'um -\' \-~'-_..‘*vu-1U!l!l! l'unl -\' \--'Tl*vu-1tm l'~l«l -\' \--'Tl*vu-1uw l '!J¢1I -\' \-.'-_lAw--r’m l '!J¢1I -\' Fungi!“ l at... -\' t..-.—u..._ . 1B INTRODUCING}! using digital processing. Radiology ’13. 2.. H. Cho (ed). Special issue on lr nuclear medicine intriimenlation. In 8. J. F. Lindcman. The re'cent history 0 he!“ ed. Baltimore: Diagnostic Nuclear Medicine, A. Goltsehalk and E. 1. Pole rl Press,1976 pp. 5—12. I I A _ 9 :83: inger Use of a gamma-ray pinhole camera lor mam-to sludles. Miturir ' ’ ' se section seans . . Edwards. Reorganiiung data from Il‘dl'thJ' '2‘ Km and R O 91975—933 (was). 11 til/Browne“ and W. H. Sweet. Localization of brain tumors. 12 (Slgiilnkoadtz J S Robertson. and W. A. Higginhotham. Positron scanner [or I locating braln‘tumorit. JRE intent. Carin. Record, Part 9:49—5h (Moll. I A physical and computational aspects ol 3~tlimcn- eternal image reconstruction. [BEE Ti'fl'i‘tt. Nitet‘. Sci. NS~21 (1974). ' - ' " ' e ‘1. A positron emission . . Ter» =ian. M. E. Phelps. E. J. Hellman. 'l-d I ransaxialiflmgmph [or nuclear imaging (Pi-SIT). Radiology [1439—98 ([975). I“? 2 H. (‘ho J. K. Chan, and I... Eriksson. Circular ring transverse axial positron H camera/for, 3-D reconstruction of radiomiclidc distribution. iiZEE Trans. Nttci. Sci. NS~2316l3—fi22(l976}. 16. F. Budiniier, S. E. Derenzo. G. 'l“. Ntu'leottics 1 1:4 0—4 5 Gnllberg, W. L. Greenberg. and l‘l, Y’i‘f'luesman. Emissietnrfimpuled tomograpr with single photon .and posuron annihilation photon emitters. J. (Iotttpttt. Assist. Tomogr. 1:13I—-14:1 (1)77). 13'. 2. H. (fire and M. R. Farukhi. Bismuth germinate (EGO) as a potential scintilla- lion detector inposilron camera. J. Nitci. Med. 18:840-844 [1977}. I 18-.‘25'. Tanalra, N. Nohara, T. Toniitaiii, M. Yamaolo. and M. Murayama. Stationary {aid-mitten emission tomography and its image reconstruction. IEEE Trims. Med. intrtg. M'liecf‘izllJEI-Ztlfi {1986). £219 R. Jaszcrak, L. Chang, N. Stein, and F, Moore. Whole body SPECT using dual, 2.5: largexfield of-vicw scintillation cameras. Phys. Med. Biol. 24“ l23—l [43 (1979). 20. R. E. Alvarez and A. Maeovslti. Energy selective reconstruction in X-ray compat- erized tomography. Phys. Med. Biol. 21:733—744 (1976). Di Chiro. R. Affliction, R. M. Kessler, et aI. Tissue signatures with dual-energy " Cr. aetiology l31:521 (1.979). 22. Z. (Em, C. M. Tsai, and G. Wilson. Study of contrast and modulation mecl‘ifihisms in X-ray/photon transverse axial transmission tomography. Phys. Med. Biol. 20:879—839 (1975). 23. E. (Z. McCullough. Factors affecting the use of quantitative information from a CT scanner. Radiology 124:9? (I977). bl. 2.. H. Cho, 1. Aim, C. Bohm, and G. Hulh. Computerized image reconstruction methods with multiple photon/X-ray transmission scanning. Phys. Med. Biol. 19:511—522 (I974). 25. L. Shepp and B. began. The Fourier reconstruction oi a head section. “5155 Tit-ans. Nucl. Sci. his-21:2! {1974). 26. P. Lauterbur. Image (urination by induced local interactions: Example employing nuclear magnetic resonance. Nature 242:190 (1973). ....-§... 3; s is; 2T. 28. 29. fill. 3|. 32. 33. 34. 35. 36. 3?. all]. 41. 42. 43. REFERENCES 19 R. Damatiian. 'I‘uutor detection by NMR .S'et‘cnrre I'll:l I51 (19”). A. Kumar. D. Welti, and R. Ernst. NMR Fourier zeugmalography. J. Magn. Reson. Illlfill “975). W. l-linshaw. P. Bottotiiley, and 0. Holland. Nature 2781722 (I917). I’. Mansfield and A. Maudsley. Fast scan proton densin imaging by NMR. J'. Phys. ii“: Sci. Jtt.i'trttttt. 9:27] [1976). 2. H. Ciro, H. S. Kim. H. B. Song, and J. (Trimmings. Fourier transform nuclear magnetic rcsonttoce tomographic imaging. i’tmr. {BEE 701152—1173 (1982). W, I-Iinshau and A. Lord. .-\n introduction to isth imaging: From the Bloch equation to the image. equation. Prue. JEEJE 71:338 U983). K. R. Erikson. F. J. Fry, and J. P. Jones. Ultrasound in medicine—A review. iEEE Timur. Sonics ix'itt'asuitit‘s SLiw211'144w-17U [lg-id). J. P. Jones and S. Let-man. Ultrasonic tissue characterization: A review. Arta Electmtticri 2M1—1]:3‘-31 (1034). J. P. Jones and R. Kovack. A computerized data analysis system for ultrasonic tissue t-liiiritt'teiizatioti. Acotii'tit'itt' foraging 9250.1.»513 [1980']. F. (3. Somtner. l... Jot-m. H. Carroll, and A. h-lztcorslt’i. Ultrasonic characterization ol‘ abdominal tissues via digital analysis of haekscuueied was-ciorms. Radiology l-lltflllmtil? (1951:]. J. M. C. [Triallet and J. P. Jones. The envelope. correlation spectrum as a means (or characterizing Iissiie structure. Ultrasound in Medicine and Biology (in press). . M. R. Milunski. (i. A. Mohr. J. E. Perez, Z. Vcred, K. A. Wear. C‘. J. Gesslcr, 13. E. Sobel. J. G. Millet, and S. A. Wickliue. Ultrasonic tissne eharacteriratiou with integrated hackscatter. (.‘ireutati‘on EGAN-603 (1989}. . l.. J. 'l"i’toni'.is, B. l'larattlai, J. l’crer. H. E. Sobel. S. A. Wickline, and J. G. Miller. Quantitative real-time imaging of myocardium hasetl on ultrasonic inte- grated haeksmttei’. iii—Elf Trims. Ultrasonic, Fi-rrmlectrz‘cs and Frequency (L'ontrol' 3bt46t‘tw4'ill ( “3891 J. P. Jones. Ultrasonic impediography and its application to tissue characteriza- tion. In Recent Admin-cs iii Ultrasound. it: Hirziiieriiirtlne, D. N. White. ed. Research Studies Press. IQTT. pp. 131—156. S. Leenirin. lmpediography equations. .‘Zt’rttts‘iltftli bringing R:Sl‘l-—525 (1950}. 8. Lemon. Ititpctliogriiphy revisited. .Jt'oustu-ril Jtnagttii: 01513—510 (1980). I). Nicholas am] A. M. Nicholas. 'l“woAdimensional diffraction scanning of normal and cancerous huniait hepatic tissue in \‘ill‘l‘l. Liltriisotitui Med. and Biol. 9133-296 (1083). . J. P. Jones. Current problems in Ultrasonic itnpediography. lo Ultrasonic Timie Characterization. M. Litizer, ct]. NBS Spec. I’uhl. 453:253—258 (1976). . L. I‘illitlilll'ls and S. Lccman. Pulse and impulse. response in [tumult tissues. .‘lt't’ifliiflfii imaging 12:459-467 0982). . L. Ferrari, .l. P. Jones. V. Gonzalez, cl :il. Acoustical imaging using the phase of the echo waveforms. Presented at the 12th International Symposium on Acousti- cal Imaging, London. [982. Also in Acoustical imaging. vol. l2, IS. A. Ash and C. R. Hill, eds. New York: Plenum. l‘lSE, pp. 635—642. ...
View Full Document

This note was uploaded on 02/27/2008 for the course BME 525 taught by Professor Singh during the Fall '07 term at USC.

Page1 / 9

BME525_ch1 - )l ‘é 3.x 21.3:i.3.5:2 3 t l “i sit so...

This preview shows document pages 1 - 9. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online