Nuclear Medicine Resources Manual

Подождите немного. Документ загружается.

CHAPTER 5. GUIDELINES FOR GENERAL IMAGING

170

Typically, a total of at least 5 million counts in the entire study will

provide sufficient statistics for quantitative and functional image processing.

5.1.3.8. Pinhole imaging

Pinhole imaging provides the spatial resolution that most closely

approaches the intrinsic limit of the camera at the expense of sensitivity. The

distance between the collimator and the patient determines both the degree of

magnification and the sensitivity (or count rate). Smaller pinhole apertures (2–

3 mm) provide better resolution but lower sensitivity. The largest pinhole in

routine clinical use should be 5 mm. For bone scintigraphy, 4 mm is commonly

used. For typical collimators with a 25 cm diameter FOV, a matrix size of 256 ×

256 or 128 × 128 is generally sufficient.

5.1.4. SPECT image acquisition

In SPECT, a similar approach to that used for planar imaging is taken to

determine appropriate matrix size, except that pixel size must be less than

reconstructed FWHM divided by 2.5 (or 3). The acquisition matrix size will

normally be 64 × 64 or 128 × 128 depending on the reconstructed resolution

and field size. The manufacturer’s processing protocols should be consulted for

compatibility with specific data acquisitions.

The number of projections is likewise determined from similar sampling

considerations. Consider a region, centred on the centre of rotation that

includes the organ of interest. Then the arc at the edge of this region, defined

by the detector position in two adjacent projections, should be approximately

equal to the defined pixel size. In general, at least 60 (64) views are used for

360º acquisition or 30 (32) views for 180º acquisition. However, 120 (128) views

should be used for high resolution studies such as those of the brain,

irrespective of the matrix size used.

Statistics play an important role in the reconstruction process and

typically can prolong imaging times. Total imaging time should not exceed

min, to minimize patient motion.

SPECT data can be acquired using step-and-shoot, continuous motion or

a hybrid technique, depending on the camera design and the type of study to be

performed. Continuous rotation will provide the most efficient image gathering

capability, especially if 120 (128) views are acquired.

5.2. NUCLEAR CARDIOLOGY

171

5.2. NUCLEAR CARDIOLOGY

5.2.1. Introduction

Nuclear cardiology is a superspecialty, in which nuclear physicians with

training in cardiology, or cardiologists with nuclear medicine training, use

nuclear imaging technology to investigate a variety of physiological and patho

logical aspects of the cardiovascular system. The major techniques used in

nuclear cardiology can be categorized as: first pass angiocardiography, multi-

gated blood pool imaging, myocardial perfusion imaging, and receptor and

metabolic imaging. The data derived from these studies can be used for

diagnosis, prognosis, treatment monitoring and assessment of viability in heart

diseases, particularly in coronary artery disease.

5.2.2. First pass angiocardiography

5.2.2.1. Principle

First pass radionuclide angiocardiography (FPRNA) is a rapid, reliable

and reproducible method of assessing cardiac function based on a few

monitored heartbeats. It involves the imaging of an intravenously injected

radionuclide bolus during its initial transit through the central circulation. A

time–activity curve is generated, and the temporal separation of the right and

left ventricular phases allows evaluation of individual ventricular function. This

is based on the assumption that thorough mixing of the tracer has occurred in

the blood pool and that the detected count rate reflects the changes in

ventricular volume during contraction and relaxation.

Left and right ventricular function assessed at rest, or during stress with

first pass imaging, gives a comprehensive evaluation of short duration changes

that may affect the ventricles. This includes evaluation of global and regional

wall motion, estimation of ejection fraction and other systolic and diastolic

parameters. Such information has proved significant in the diagnosis, prognosis,

decision making and management of certain clinical problems such as coronary

artery disease and chronic obstructive lung disease, as well as congenital and

valvular heart disease.

CHAPTER 5. GUIDELINES FOR GENERAL IMAGING

172

5.2.2.2. Clinical indications

(a) Coronary artery disease

Rest and exercise FPRNA has been used extensively in diagnosing and

managing patients with known or suspected coronary artery disease, as well as

with complications of acute myocardial infarction.

Wall motion abnormalities, changes in end systolic volume and changes in

diastolic filling rate are suggestive of ischaemia and the presence of coronary

artery disease. Evaluation of the left ventricular regional wall motion is,

however, limited since the FPRNA is done one projection at a time. The

anterolateral, apical and inferoseptal walls are visible in the anterior view while

the right anterior oblique (RAO) view provides images of the anterior, apical

and inferior walls. Regional wall motion in the RAO first pass study correlates

well with contrast angiocardiography. Correlation is likewise good between the

left ventricular ejection fraction (LVEF) calculated using FPRNA and the

values obtained by contrast and equilibrium radionuclide ventriculography. For

the measurement of right ventricular ejection fraction (RVEF), FPRNA is

considered the method of choice. The technique allows images to be acquired

in the RAO view, giving a maximal separation of the right atrium from the right

ventricle. The tracer bolus might not, however, mix completely with the right

atrial blood prior to entering the right ventricle and may exit without mixing

completely with apical blood, giving rise to potential sources of errors.

(b) Congenital heart disease

Detection and quantification of left-to-right shunts are possible using

FPRNA. The presence of a shunt is confirmed by simultaneous tracer

appearance in the right and left ventricles. In a normal study, the left ventricle

is free of any right ventricular activity. Quantitation of a left-to-right shunt is

dependent on the quality of the bolus injected. A delayed or fragmented bolus

may affect the shape of the pulmonary curve generated, which should be

monoexponential, even in the absence of a shunt. Shunting separates the

pulmonary activity curve into two components, which are proportional to the

systemic and shunt flows, respectively, giving an index of the severity of the

shunt.

Assessment of valvular insufficiency is possible with resting FPRNA.

Studies showing prolonged tracer transit through the left side of the heart may

5.2. NUCLEAR CARDIOLOGY

173

reflect mitral or aortic insufficiency. From the pulmonary and left ventricular

time–activity curves, the degree of regurgitation may be calculated and

quantified. Data from FPRNA have been shown to correlate with measure

ments obtained by catheterization. Resting studies performed serially can be

helpful in monitoring the severity of the valvular insufficiency and in deciding

when valve replacement is necessary.

5.2.2.3. Radiopharmaceuticals

The ideal radionuclide as a first pass imaging agent must remain intravas-

cular as it moves through the central circulation. It should also be safe for

application in large doses in order to generate the necessary high count rates.

(a) Technetium-99m agents

The high specific activity of

99m

Tc makes it suitable as a first pass agent.

For multiple or sequential studies,

99m

Tc diethylene triamine pentaacetic acid

(DTPA) is preferred to

99m

Tc-pertechnetate. DTPA has rapid renal excretion,

making possible a repeat injection 20 min later. Technetium-99m pertechnetate

can be used when a single assessment of ventricular function is needed. The

usual dose for

99m

Tc agents is 925 MBq and a maximum of three injections,

370

MBq (10 mCi) each, can be given (Table 5.1).

Other technetium based compounds such as sestamibi and tetrofosmin

are also suitable. First pass imaging can be performed upon injection of the

tracer during peak exercise, thus combining information on regional and global

ventricular function as well as myocardial perfusion in one setting.

(b) Short lived radioisotopes

The need for first pass studies to be performed repeatedly in a short

period of time presents some restrictions with

99m

Tc agents. The half-life of 6 h

and varying biological clearance times limit the number of acquisitions that can

be done in a given period. In order to reduce the patient’s radiation exposure

and allow for a greater number of studies to be performed, radionuclides with

half-lives in terms of seconds or minutes would be ideal.

Tantalum-178 produces suboptimal results when used with standard

gamma cameras because of its low energy; more satisfactory results have been

reported with a multiwire proportional gamma camera. The short half-life of

191m

Ir makes it suitable for paediatric patients. Gold-195m is ideal for adult

patients, and the calculated ejection fraction correlates well with that obtained

using

99m

Tc agents. Evaluation of myocardial perfusion could possibly be done

CHAPTER 5. GUIDELINES FOR GENERAL IMAGING

174

with

195m

Au in combination with ventricular function assessment. Table 5.2

summarizes the physical characteristics of the above mentioned short lived

radioisotopes.

5.2.2.4. Equipment

Monitoring of the radionuclide bolus during its transit through the central

circulation requires a gamma camera with a high count rate capability, a high

sensitivity collimator, a computer software capable of processing high temporal

resolution data and a small matrix acquisition, as well as an ECG gating device.

A bicycle ergometer is an additional requirement for first pass studies during

exercise.

More than 200 000 counts/s are needed to achieve an adequate image

quality. The use of cameras with low count rate capabilities leads to an

inaccurate measurement of ejection fraction and assessment of wall motion.

Originally, only multicrystal gamma cameras could record such high counts,

although with some loss of spatial resolution. Newer generations of multi

crystal cameras can now acquire the same range of counts with enhanced

energy and spatial resolutions. Modern single crystal cameras are also capable

of achieving rates of up to 200 000 counts/s, as opposed to older cameras with

rates of only up to 60 000 counts/s.

The choice of collimator depends on the objective of the study and the

dose to be injected. Computer software should allow acquisitions to be

performed with 64

¥ 64 or smaller matrices.

An ECG signal is unnecessary for high count rate studies, since there are

enough counts to distinguish between end-diastolic and end-systolic frames. It

TABLE 5.1. FIRST PASS STANDARD Tc-99m DOSES

Type of study Rest Exercise

LV / RV

function –

multicrystal camera

370–925 MBq (10–25 mCi)

(0.3–0.5 mL)

925 MBq (25 mCi)

(0.3–0.5 mL)

LV function –

single crystal camera

925 MBq (25 mCi)

(0.3–0.5 mL)

925 MBq (25 mCi)

(0.3–0.5 mL)

RV function –

single crystal camera

740–925 MBq (20–25 mCi)

(0.3–0.5 mL)

740–925 MBq (20–25 mCi)

(0.3–0.5 mL)

Shunt study 370–555 MBq (10–15 mCi)

(0.3–0.5 mL)

LV, left ventricular; RV, right ventricular.

5.2. NUCLEAR CARDIOLOGY

175

may be more helpful with single crystal camera acquisitions to facilitate data

processing of low count rate studies.

5.2.2.5. Procedure

(a) Tracer injection

First pass studies require the injection of a small volume of radionuclide

bolus. The quality of the study largely depends on the integrity of the bolus.

Large proximal veins must be used as injection sites, since smaller, peripheral

veins may cause bolus fragmentation. The injection parameters appropriate to

the various kinds of study are listed in Table 5.3.

For left ventricular evaluation or shunt studies, it is important that the

bolus arrive in the heart as a single front. Rapid injection of the radionuclide

and a 10–20 mL saline flush (within 2–3 s) is necessary. In right ventricular

studies, since the bolus reaches the right ventricle without significant

dispersion, an antecubital vein is preferred since the use of the external jugular

vein may result in too rapid transit of the bolus through the chamber. A slower

bolus is preferred to increase the number of beats available for analysis; the

saline flush may be then infused without interruption for 3–4 min.

(b) Radionuclide dose

The appropriate dose depends upon the sensitivity of the gamma camera,

the number of times the study will be performed and the specific radionuclide

used. Recommended doses for

99m

Tc agents are listed in Table 5.1. Short lived

radionuclides have been given at doses as high as 50 mCi. The acquisition

parameters for FPRNA are given in Table 5.4.

TABLE 5.2. CHARACTERISTICS OF SHORT LIVED RADIOISOTOPES

Radioisotope Half-life Photon energy (keV)

Ta-178 9.3 min 55–65

Ir-191m 4.9 s 129

Au-195m 30.5 s 262

CHAPTER 5. GUIDELINES FOR GENERAL IMAGING

176

Proper positioning of the patient must be verified prior to the start of the

study. A radioactive source or dose syringe may be used to check areas of

interest in the FOV and allow identification of the lungs.

Temporal separation of the cardiac chambers allows the study to be

acquired in any view, although the RAO and anterior projections are most

commonly used. The shallow RAO view — best for direct comparison with

contrast angiography — separates the atria from the ventricles and the left

ventricle from the descending aorta. The 30

RAO view enhances the

separation of the right chambers, making it the ideal view for right ventricular

assessment.

The upright straight anterior view is best for exercise studies since the

chest is stabilized against the detector. The pulmonary background is also

reduced, enhancing study quality. The descending aorta and the basal portion

of the inferoseptal wall may, however, overlap with the left atrium and basal

portion of the left ventricle. The left anterior oblique view is useful when the

circumflex artery territory is in question, but may result in underestimation of

the LVEF.

(d) Frame rates

A standard frame time of 25 ms/frame can be applied for RV and LV

function studies regardless of the type of camera used, but theoretically the

frame time should be adjusted according to the heart rate. Fifty ms/frame is

adequate at heart rates lower than 80 beats per minute decreasing to 10–20 ms/

frame for faster heart rates, especially if diastolic function is of interest. Two

thousand frames are sufficient to encompass the entire left ventricular phase.

TABLE 5.3. INJECTION PARAMETERS

function studies RV

function studies Shunt studies

Site Medial antecubital or

external jugular vein

Antecubital vein Antecubital or external

jugular vein

Cannula 18–20 gauge 18–20 gauge 18–20 gauge

Rate Rapid (FWHM < 1 s) Slow (FWHM 2–3 s) Rapid (FWHM < 1 s)

LV, left ventricular.

RV, right ventricular.

5.2. NUCLEAR CARDIOLOGY

177

Frame rates are not as essential in a shunt study since data analysis uses curves

of lower temporal resolution.

(e) Stress protocol

Bicycle ergometry is the method of choice for exercise studies. Although

supine bicycle exercise results have been shown to correlate with catheteri

zation, upright bicycles are more often used since they minimize chest motion

and are better tolerated by patients. Any graded exercise protocol is acceptable

and no time is required to stabilize the heart rate. Exercise should be continued

until the bolus clears from the left ventricle.

(f) Data processing

Processing of a first pass radionuclide study can be divided into four

major steps:

(1) Generation of an initial time–activity curve over the cardiac region;

(2) Beat selection;

(3) Background subtraction;

(4) Creation of the final representative cardiac cycle.

The initial time–activity curve is important for quality control. It permits

inspection of the separation of the right and left ventricular phases, allows the

estimation of the peak count achieved, and detects the presence of irregular

beats. An ROI is drawn around the ventricle, grouping frames from the raw

data. This is used to generate a time–activity curve from which an initial repre

sentative cycle is created and used to draw separated end-diastolic and end-

systolic ROIs.

TABLE 5.4. ACQUISITION PARAMETERS FOR FPRNA

Parameter LV

function RV

function Shunt study

Position Upright Upright Upright

Angle Anterior 20–30

RAO Anterior

Frame time (ms) 25 25 50

Total frames 1500–2000 1500–2000 2000

LV, left ventricular.

RV, right ventricular.

CHAPTER 5. GUIDELINES FOR GENERAL IMAGING

178

Once the left ventricular ROI has been identified, cardiac cycles to be

included in the final analysis may be selected from the ventricular time–

activity curve. The cycles before and after the beat with the maximum

number of counts are selected. Premature ventricular beats and post-

extrasystolic beats should be excluded. Beats whose end-diastolic counts are

below 50% of the maximum end-diastolic count should also be omitted if

they do not preclude a statistically adequate representative cycle. Only beats

around the peak of the time–activity curve (80% or more of maximum

activity) are to be used. This leaves one or two beats during the right

ventricular phase and four to five beats during the left ventricular phase

available for analysis. Averaging of several individual beats can also be done

to form a summed representative cycle.

Background subtraction can be performed using several methods. The

most accurate appears to be the lung frame method. Image counts in the ROI

just prior to the appearance of tracer activity in the left ventricle are chosen as

background counts and used to correct the left ventricular phase. Background

correction is crucial for the LVEF, and any variations in background counts can

lead to changes in the calculated ejection fraction, volumes and wall motion.

Once the background has been corrected, initial ROIs are adjusted and

the final ROIs are used to regenerate the time–activity curve from which the

final representative cycle is created from the previously selected beats. This

cycle may be displayed in a cine-loop for analysis of regional wall motion. All

quantitative data are also derived from this cycle.

The LVEF and RVEF are calculated from the end-diastolic (ED) and

end-systolic (ES) counts as follows: (ED counts – ES counts)/ED counts. The

systolic emptying rates and diastolic filling rates are calculated with

appropriate software using a Fourier filter applied to the representative cycle

and taking the first derivative of the filtered curve. Left ventricular end-

diastolic volume may be measured using the geometric or count proportional

method. The geometric method measures the area of the left ventricle and the

length of the major axis in pixels. Converted to centimetres, the pixels are used

to calculate the volume. In the count proportional method, volume is derived

from the total counts and the counts in the hottest pixel in the left ventricle.

This method requires validation for each laboratory.

5.2.2.6. Interpretation

The radionuclide bolus appears sequentially in the superior vena cava,

right atrium, right ventricle, pulmonary circulation, left side of the heart and

aorta. Any changes in this pattern would suggest the presence of a congenital

abnormality. Delayed tracer transit through the right side of the heart suggests

5.2. NUCLEAR CARDIOLOGY

179

pulmonary hypertension, tricuspid or pulmonary valve insufficiency or a left-

to-right shunt. Delayed tracer transit on the left side of the heart would suggest

mitral or aortic insufficiency.

Regional wall motion is analysed by superimposing the end-diastolic

outline against the end-systolic image or by viewing the representative cycle in

cine-mode. However, it has to be noted that since the study was acquired in

only one projection, regional wall motion abnormalities may be difficult to

identify in overlapping segments. Ischaemic responses applicable to the

diagnosis of coronary artery disease are typically a new onset of a regional wall

motion abnormality or a worsening of a previous one, an increase in the end-

systolic volume and alterations in diastolic filling parameters.

BIBLIOGRAPHY TO SECTION 5.2.2

BAILET, G.Y., et al., Simultaneous technetium-99m MIBI angiography and myocardial

perfusion imaging, J. Nucl. Med. 30 (1989) 38.

FOLLAND, E.D., et al., The radionuclide ejection fraction: A comparison of three

radionuclide techniques with contrast angiocardiography, J. Nucl. Med. 18 (1977)

1159.

GAL, R., et al., High count rate first-pass radionuclide angiography using a digital

camera, J. Nucl. Med. 27 (1986) 198–206.

GARCIA, E.V. (Ed.), Imaging guidelines for nuclear cardiology procedures (Part 1), J.

Nucl. Cardiol. 3 (1996) 999.

GERSON, M.C. (Ed.), Cardiac Nuclear Medicine, 3rd edn, McGraw-Hill, New York

(1997).

MENA, I., NARAHARA, K.A., DE JONG, R., MAUBLANT, J., Gold-195m, an ultra-

short-lived generator produced radionuclide: Clinical application in sequential first-pass

ventriculography, J. Nucl. Med. 24 (1983) 139–144.

PORT, S.C., Recent advances in first-pass radionuclide angiography, Cardiol. Clin. 12

(1994) 359–372.

5.2.3. Equilibrium radionuclide angiography

5.2.3.1. Principle

Equilibrium radionuclide angiography (ERNA) is a non-invasive means

of quantitatively assessing cardiac function. It has been demonstrated to be a