Indian Journal of Biochemistry & Biophysics Vol. 36, October 1999, pp. 34 1-347

Review

Magnetic resonance imaging: Bioeffects and safety concerns

N R Jagannathan

Dcpartmcnt or N.M.R., All India Institutc or Medical Scienccs, Ansari Nagar, New Dclhi 110029

Magnetic resonance imaging (MRI) is the state-of-the-art noninvasive imag ing modality in clinical di agnosis. During MRI exami nati on, the patient is exposed to three different forms of electromagnetic rad iati on: (i) a stati c magnetic field , (ii) grad ient magnetic fields , and (i ii) radiofrequency (RF) fi elds. Each of these may cause significant adverse bioeffects if applied at sufficiently high exposure levels. This arti cle desc ribos in so me detail the areas of health concern for both the patient and the health practitioner with respect to the li se of clinical MRI, in add ition to describing the potential bioeffects of elec tro magneti c radiations lI sed in thi s sophisti cated imag ing modalit y.

Introduction Magnetic resonance imaging (M RI) is presentl y an

accepted and widely used clinical diagnostic procedure. This nonin vas ive anatomic/pathol ogic imagi ng technique utili zes magneti c fields and radiofrequency (RF) energy to manipulate atomic nuclei present in human tissues I . The most commonly used nucleus in MRI is hyd rogen (proton) because of its relat ively strong signal and 99% natural abundance in bi ologic ti ssues. Most of these protons are assoc iated wi th water and, to a lesser extent , fat. When placed in a strong magnetic fie ld , these protons behave like tiny bar magnets and tend to orient in a direction along the magnetic field. An RF field applied at ri ght angles wi II change the orientati on of the protons within the magnetic fi eld . Transfer of energy occurs as a "resonance" specifi c to both the nuclei (i.e., protons) studi ed and the precise magneti c field surrounding it. After the terminati on of the RF exc itati on, the protons are free to return to their normal orientat ion in the field (relaxation) by givi ng up the energy they have absorbed which is detected by a receiver coi l as a signal (free inducti on decay, FID). Through the use of magnetic fi eld gradients and RF pulses, the location of the nucl ei in 3-D space, as well as the information about its surrounding environment, can be encoded into the returned signal l

.

The excitation and detection process is repeated several times to uniquely encode each part of the object. Computer processing of the set of signal s (FIDs) provides the final image(s) of the spatial

Fax: 9 1-0 I 1-686 2663 or 652 104 1 Tel : 9 1-0 I 1-6850533 E-mail: nrj @mcdinst.ernet.i n

di stribution of the nuclei within the object. Thus, it is clear that MRI represents one of the most sophi sticated technique with a compos ite of diverse technological components such as a powe rfu I static magnetic field, radiofrequency (RF) fi elds, rapid ly switching local magnetic fi eld s ca ll ed gradi ents , vast amounts of cryogens and powerful computers. Thus , during the MRI in ves ti ga ti onal proced ure, the pat ient is exposed to three different forms of electromagnetic radiation: (i) a static magnet ic fi eld , (i i) gradient magnetic fields, and (iii ) radiofrequency fi eld s. Each of these, individually or co llect ive ly, may cause significant bioeffects, if app li ed at sufficientl y hi gh exposure levels .

Various safety considerati ons of clini cal MRL ha ve been investi gated by a number of groupS2. I 0

conclusive ev idence of significant hea lth hazard due to ex posure of these radiations in human subjects has been reported by these investi gati ons". In additi on to the bioeffects related to ex posure to elec tromagneti c fields used in thi s imaging modality, thi s arti cle describes in some detail the areas of hea lth concern for both the pat ient and the hea lth practiti oner. However, before proceeding furthe r, it may be menti oned emphatically that MRI is a noninvasi ve technique with no ioni zing radiation and therefore poses no radiation hazard to hea lth.

Static magnetic fields The most important component of the M R scanner

is the magnet which produces a constant static magnetic field. The magnet in a typ ica l MRI scanner in use today produces a field between 0.2 T and 2.0 T in terms of the familiar magnet ic field units of Tesla (T), IT = 10,000 gauss. No sc ienti fic ev idence ex ists

342 INDIAN 1. BIOCI-IEM. BIQPHYS., VOL. 36, OCTOB ER 1999

to show that magnetic fi elds of above range produce harmful effects in mammals . Hence, it is totally safe while the patient is in side the MRI magnet. MRI scanners, presently in use, are ' shielded' whereby the above fi eld strength s are mostl y confined inside the MRI suite, and only very weak res idual or ' frin ge ' fields (- 0.5 mT or 5 gauss) pervade a small peri meter around the MRI magnet.

The attrac ti ve force of the magnet on ferromagnetic objects has the potenti al to adversely affect the safe ty of the patient as well as th,e MR personnel. Such attractive forces include both translati onal (i .e., attrac ti on by the magnet) and torsional (attractive force causing the long axis of the ferrous objects to align with the magneti c fi eld). ' Ferromagnetics ' are materi als that contain iron and arc strongly attracted by a magnet. In addi tion , metal s li ke ni ckel and cobalt are also strongly ferromagneti c. Metals such as copper, aluminium, sil ve r, and go ld are non­ferromagneti c. It is essential to screen objects both in and arou nd the patient as well as th ose within the patient. In the former group, objects ranging from forklifts, wheel chairs and isolets, to pens, pencils, and paper clips have been reported as having been inadvertently attracted to an MR scanner. Due to attractive forces produced by a whole body MR scanner, even fairly li ght objects may cause significa nt injury. For exa mp le, a paper clip in a I .ST MR system reaches a terminal ve loc ity of approxi mately 40 mph. It is our experience that adequ ate restriction of access to the magnet, continual educati on and screening of M R personnel, and appropri ate patient screening helps to prevent such incidents . At our site, the patient screening procedure consists of (a) a carefu l wri tt en query, (b) scanning the pati ent top-to-t oe with a hand held metal detector (after the removal of keys, coins, hairpins, credit cards, etc., and (c) checking any previ ous radiological filmslimages of the pati ent that may show implanted objects.

Another area of concern is the magneticall y sensitive equipments/objects, the function of which may be adversely affected by the magneti c fie ld and MR imaging would be contraindicated even more strongly. The most common of these items is the cardi ac pace makers as we ll as other implantable electronic devices such as intraocul ar metal, Coh lear implants, ferrous intracranial aneurys m clips and sharpnel in critical locati ons. It is found that field strengt hs greater than 10mT are suffi cient to close the magneti c reed switch of the pacemaker, sw itching it

to an asynchronous mode . Although thi s is usuall y not a threat to the patient for short peri ods, the poss ibility that some harm can result docs cx ist. For examp le, thc lithium battery which powers the pacemak er may al so become di slodged in magneti c fi e lds. III view or thi s the public access to areas with fi elds of 0.5 IllT and greater must be posted with warnings. or access contro ll ed.

In additi on, the 0.5 mT I'r inge I'ie ld Illay al so al'rcc t the working of other vit al medi cal dev ices such as patient monitoring equipmcnt , respirators, col our television monitors, magneti c tapes, fl oppy di sks, etc . Necessary precauti onary mcasures ha ve to be taken for proper functi oning of these equi pmcnt s. Attenti on must also be given to the placement or suitable anaesthes ia close to the MRl suitc, necess it y of shi elding them magneti call y, etc. should be assessed from MRI manufacturers bel'ore in stalla tion.

Exposure to static magnct ie fi c lcl docs no t alter skin and body temperatures in humans accordin g to a recent study at 1.5 T performed usin g ; 1 spec ia l fluoroptic thermometry-' . Therel'o re, sk in and body temperatures on human subjects arc be li cved tu be unaffected by exposure to stati c magne ti c fi c ld s or upto 1.5 Tesla4

.5

.

In additi on to the harml'ul efrcch I'rum mctal objects and ferromagneti c prosthes is being attracted by the magnetic fi eld , there are several bi oph ysica l mechanisms whereby static magneti c fi elds mi ght influence biological processes or organi sati onal behaviour. (i) Changes in enzyme kineti cs by conformational changes, by perturbati on or free radica l mobility, or by qucnchin g superconductiv ity thought to be present in biomoleeular systems . A wide variety of experiments at field s upt o 45 T ha ve not shown any effect on enzyme systems2

. ( ii) Orientati on changes of macromolecules and li vin g ce ll subce llular components, leading to changcs in chemi cal kinctics and membrane permeability. ( iii ) Di storti on or ion currents lead ing to a reducti on of nerve conducti on velocity. Theoreticall y, e lectrical impul se conducti on in nerve tissue may be affected by exposure to stati c magnetic fi elds. At present, exposure to sUlti e magneti c fi elds of up to 2.0 T do not appea r to significantly influence bioe lectri c properti es of neurons in humans6

-R

. (iv) lnduccd e lcctric potentia ls and currents in the cardi o-vascular systeill . The induced voltage for a field at ri ght angles to the fl ow direction is given by

JAGANNATHAN : MAGNETIC RESONANCE IM AG ING: BIOEFFECTS AND SAFETY CO CERNS ~-n

F=q(vxB) __ . ( I )

where q is the charge, v is the ve locity and B is the magnetic fi eld. Therefore,

Y=vBd ... (2)

Here Y is the voltage, and d is the diameter of a vessel (meters) through wh ich there is a uniform velocity of v in m/sec. The maximum flow velocity in the cardiovascular system occurs in the large arteries . Thus the ampl itude of the e lectri c potential is greatest there. The experimentall y observed induced potential in a magnet occurs during the T-wave of the electrocardi ograph (EKG). The threshold fi eld for a percept ible T-wave change is below 0.3 Tes la'} The T-wave abnormality in creas~s with field strength , however, thi s induced vo ltage does not result in any physiologica l changes . (v) Magnetohyd rodynami c effects leadi ng to blood ve loc ity decrease and blood pressure elevati ons. The electrodynamic intent tion of stationary magneti c fi elds wi th fl owing, electri call y conductive fluid s has a retardati on effect as fluid motion which can result in an increase in the blood pressure needed to deli ver a given v~lume of blood to peripheral ti ssues. These effects are not significant either in theory or from experiments upto 1.5 T. However, a significant blood pressure increase of 28% in the aorta and venocava is theoretica ll y predicted for a field of lOT. There shou Id be no effect on the blood pressure in the rest of the vessels in the vasculature since the effect is proportional to the square of the vessel radius . Several experiments carried out revealed that there is no significant magnetohydrodynamic pressure increase even at 10 Tesla2

•9

.

In summary, there is no conc lusive evidence of irreversible or hazardous bi ologica l effects related to acute , short-term ex pos ures of humans to stati c magnetic fi elds of strengths upto 2.0 T . However, there are several 3.0 T and 4.0 T wh ole bod y MR systems now operationa l at various cen ters wor ld wide. A pre l i mi nary st ud y has i nd icated th at workers and vo lunteers ex posed tu 4 .0 T MR system ha ve experienced vlrtlgo, nausea, headaches, a metallic taste in their mouths, and magneto- phosphoe nes lo (w hi ch are visua l flash es). Therefore, more research is required to stud y the mechanis ms responsibl e for these bioe ffects and to determin e the poss ible means , if any, to counter balance them .

Cryogens

All superconduct ive MR systems in c linical use today util ize liqu id heliuIll. Liquid heliul1l will ac hieve gaseous stat e (' boil ofT) ,It - -268 .9 C (.+.22 K). If due to so me reason the he lium cryost;lt warms up, it deve lops ' hot spots' and there \vill be an alarmingly fast boil off of helium. In view of thi s, an enormous pressure bui lds up and the excess ga~ has to be vented out of the building through proper outle ts. On the other hand, if thi s safety valve frosts up and fail s to work, a seal in the cryosta t would rupture and vas t quantities of the cryogenic gas wou ld fill the MR I suite and thi s phenomenon is kn own as "magnct quench". Asphyxiation and frostb ite are poss ib le if a person is exposed to heliu m vapour for a long ti meX. Recentl y, better cryostat des ign and insulati on havc al lowed the use of only liq ui d hcli uI11 in man y of the newer supercond uct ing magnets. However, in the older vers ions of superconducti ng M R I magncts, bot h liquid ni trogen and li qui d h ~ liUI11 arc used . Liq uid nitrogen within the cryostat acts as a bufl er het\\cc n the liquid helium and the outside atillos phere, boiling off at 77 .3 K. In the event of an accidental re l ea ~c of liquid nitrogen into the amb ient atmos phere of the imaging room, there is a poss ihili ty for fros tbit e, similar to that encountered wi th the gaseous helium release . A pure nitrogen cnvironment is hazardo us and unconsciousness generall y result s as ear ly as :) to 10 seconds after ex posurex. Pati ents and hea lt h personnel should be evac uated from the area as soon as it is recognized that nit roge Il or helium gas is be ing re leased into the MRI scan room, and they should not return unti I appropri ate correct i ve meas u res have bee n taken to clear the gas from the roomx.

In general , cryogens prese nt a potential co nccrn in clini cal MRI despi te an overwhelming ly safe record over the 20 years of clini ca l serv icex

. AIl ox yge n monitor with an audible alarm, si tu ated at an appropriate height wi th in eac h i magi IIg rOOIll, should be a mandatory minimum sa fety measure for al l sites .

Time-varying (gradient) magneti c fields

During MRI procedure, a humaIl body is ex posed to rapi'd vari at ions of magneti c fi elds du e to transient :.Jpplication of time vary in g grad ieIlt magnctic fi e ld s during the Imaging sequence to obtain th rce­dimensional spat ia l information. These grad ie llt magnetic fields can induce electri cal f ie lds ami currents in conducti ve media (including bi ological tiss ues) according to Faraday's law of inducti on. The

344 INDI AN 1. BIOCHEM. B10PHYS., VOL. 36, OCTOBER 1999

interaction between gradient magnetic fi elds and biologic ti ssue is dependent on the fundamental fi eld frequency, the maximum flux density, the average flu x density , the presence of harmonic frequencies, the waveform characteristics of the signal, the polarity of the signal, the current distributi on in the body , and the electri ca l properties and sensiti vity of the particular cell membrane6

-8

. In animals and human subj ects, the induced current is proportional to the conductivity of the biologic tissue and the rate of h f h . t'l d . 78 II 12 C ange 0 t e magnetIc ux enslty " . .

... (3)

h J . h d " A .1 dB . T I were IS t e current enslty 111 m -, - 111 sec', dt

r is in meters and J the ti ssue conducti vity in Sm·1

(Siemens). SUPPDse we assume an effective radius of dB I

0. 1 m ( 10 cm) for the head, - of I T sec- , and (J = dt

0.2 Sm-I, then

J = 1O-2Nm2 or I~Ncm2.

Thus, changes in magnetic fi elds of 4 Tlsec can induce current densities of 4 ~AJcm2 , and if thi s is applied for a sufficient time in one direction, nonthermal biologic effects (not necessarily harmful) can be ex pected. The current density will be enhanced at hi gher frequencies and magnetic flu x densities will be further accentuated by a larger ti ssue radius with a greater ti ssue conducti vity. Moreover, di fferences in ti ssue types also affect the current density and its path : ti ssues with low conducti vity (e.g. adipose and bone) will change the pattern of the induced current.

As mentioned earlier, bioe ffects of induced currents are caused either by the power depos ited by the induced currents (thermal effec ts) or by direct effects of the current (nonthermal effects). Thermal effects due to switched gradi ents used in MRI are negli gible and are not believed to be clinically significant7

.8.L1

. The poss ible nonthermal effects of induced currents incl ude (a) stimul ati on of nerve or musc le cell s, (b) stimulati on of vi sual fl ash sensations, (c) bone healing, (d) electro shock anesthesia (therapy), (e) increased bra in mannitol space, (f) induction of ventricular fibrill at ion, and (g)

' 1 . . 16. 8 111 1 Th h hid epl eptogenlc potentI a "- . e t res 0 currents required for nerve stimulati on and ventricul ar fibrill ati on are known to be much higher than the es timated current densiti es that will be induced under

routine clinical MRI conditions(,·X.II.11. lnfac t, stimulation of muscles by means of time varying magnetic fields requires current densities of nearl y 10 times greater th an that used in MRI stratl:gies. Pulsed fi elds of the order of 10~ T/sec have bee n used to stimulate peripheral nerve trac ts .

In additi on, a visua l phenome no n may be produced if the pati ent moves hi s head rapidl y in a magnetic fi eld above about 10 mT. T hi s phenomenon, a visua l sensati on of fl as hin g li ght s kn own as magnetophosphenes, is not th ought to affect vi sion permanentl y and ceases immedi ate ly foll owing magneti c fi e ld ex pos u re I~. A I th ough there ha ve been no reported cases 0 f magnetophosphenes for fi e ld s of 2 T or less, magnetophosphenes ha ve been repo rted In vo lunteers workin g in and around a 4.0 T MR research system lO

, In additi on, a metalli c taste and symptoms of verti go also see med to be reproducible are be ing reported at th ese 4.0 T

10 systems , The rates of change of gradi ent fi e ld strength

used in MRI are of the same order of magnitude as the changing fi e ld compon nt of ext reme ly low frequeilcy (ELF) e lectromagnet ic fie lds used in th e normal domesti c transm iss ion an d d istrib uti on e lectric lines. Even th ough there has been no stud y to conc lusive ly prove th e carc inoge ni c effec ts from ex posure to ti me-vary ing magneti c fil: lds of va ri ous intensiti es and durati ons, severa l repo rts suggest th at an assoc iati on betwee n th l: two is stil l

I 'bl I S· 17 P ausl e ' . In an actual MRI sequence, grad ien l changes of

onl y the order of a few tes la per second are used during 20 IT clinica l imaging tec hniq ues. Some fas t imaging schemes may use larger induced current densities. In newer version of MRI scanne rs, a spec ial fast-imaging scan techni que ca lled Ec ho Planar Imaging (EPI) is offered as a standard protoco l. In EPI, the data is co llected in multi ple 'v iews' (or lines in k-space) of 20 imaging plane usin g a' single shot' excitati on. This methodology, demands in turn that gradient amplitudes as large as 25 mT/m , and ' slew rates' (raise time of gradi ent to its maxi mum value) close to 65 Tlmls, These are the crucial upper lim its before peripheral nerve and musc le stimu la ti on effec ts are experienced by humans. It is therefore mandatory that manufacturers of EPr MR equipment do not trade off these safety limits III favour of ' better' perfo rmance, either in imaging speed or in Image resolution,

JAGANNATHAN: MAGNETIC RESONANCE IMAGING: BIOEFFECTS AND SAFETY CO CERNS 345

Radiofrequency (RF) effects

The consequence of RF exposure of the levels used in MRJ, is the gene ration of heat in the ti ssue as a result of res isti ve losses. Therefore, the main bi oeffects associated with exposure to RF energy are re lated to the thermogenic qualities of thi s

. f' Id6-8 11 18-?? TI f h t e lectromagnetic Ie . -, --. le amount 0 ea or energy produced is a fu nction of the frequency , RF power, durati on of the exposure and coupling between the RF co il and the subj ect. The important biol ogical factors are: conductivity , specific heat of the ti ssue, size of the current loop, ci rcu lation and the abi lity o f the body to lose heat. The measure of energy or heat delivered to tissue is the power deposited times the duration of ex posure. The absorbed power per unit of mass of ti ssue is known as the spec ific absorpti on rate (SAR) which is expressed in units of Watt per kg (W/kg) . The energy per time (power) imparted to charged particles by an E field is E x J . Thus, the average powe r absorbed power per unit mass of the

induced ave rage e lectric fi e ld E l.fi IS

E crE I E 2cr SAR=-x-x-- = - ... (4)

.fi .fi mass 2M

where the energy absorbed (heat) IS the product of SAR and time.

Using f1uoroptic thermometry probes it was demonstrated that human subjects exposed to MRI and SAR levels upto 4.0 W/kg have no stati stically significant increase in body temperature and elevations in skin temperature, and they are not believed to be clinically hazardous22

. . These results imply that the suggested exposure level of 0.4 W/kg for RF energy using MRI is too conservative for individuals with normal thermoregulatory function2

., . It is important to point out that several protection mechanisms inbuilt in the MR scanner are programmed such that the above SAR limits are always obeyed .

Human organs such as testis and eye have reduced capabilities for heat diss ipation and are particularly primary sites of potential hannful effects if RF energy exposures during MRI are excess ive. One can thus calculate the temperature e levation by using the specific heat of tissue which relates the temperature change to heat absorbed. The expected theoretical temperature rise, ~t, using a soft tissue specific heat of 0.83 k cal/kg! °C (3.5 kJ/kgfC) assuming no heat loss is given by

~tCC) = SAR xTime 3.5

... (5)

where SAR is in W/kg and time in seconds. Studies carried out to measure sc rota l skin

temperatures in vo lunteer subj ects unde rgo ing MR I at a who le-body ave raged SAR of 1. 1 W/kg, re veal ed that the largest c hange in sc rota l sk in tempe rature recorded was 34.2°C. These te mpe rature c hanges were below the threshold known to impair testi cula r function 24

. Di ss ipat ion of heat from the eye is slow and ine fficient because of lack o f vascularizati on. Const~nt exposures of RF e nergy to the eyes or heads of laboratory anima ls have been de mo nstrated to be cataractogenic as a result of the therma l d isrupt io n of ocu lar tissues l 8

. Cornea l temperatures ha ve been measured in patients undergo ing MRI or the brai n using a send receive head coil at local SAR s upto 3 .1 W/kg26

. The largest cornea l tempe rature change was 1.8°C and the highest te mpe rature measured was 34°C. Animal mode l studies have demonstrated cataractogensis between 4 1 ° to SSOC for acute, nearfie ld exposure. There fo re, it does no t appear that c linica l MRI using a head co il has the potential to cause thermal damage in ocular ti ssue2s

In addition to g lobal temperature ri se, there is some concern that "hot spots" could occur. These are caused by an uneven di stributi on of RF power and in view of thi s e lec tric current path ways become narrowed in ti ssue with little vascul arity. The presence of fore ign materi als such as fe rromagnet ic prostheses comp licates the estimate of local heat ing. Moreover, exposed wires o r metal conductors shoultl never touch any part of the pati ent , and no part of the patient's body should become an induced cu rrent loop,e.g. ECG monitoring. These may induce RF burns .

RF shielding is important in many MRJ suites for two reasons: (a) protection o f the environment from the RF fields produced by MRI transmitters, and (b) protection of MRI from external RF sou rces such as TV transmitters, radio broadcasts , e mergency rad io services and pagers . This ' Faraday Cage ', as it is called, is made of thin copper foi I sheet ' walls' grounded at one point. The cage has at one place a door to allow entry of patient and staff and another place to provide ' a filter plate' ent ry for e lectrica l wires, anaesthesia gases, etc .

Acoustic noise

Rapidly-switching magnetic field gradie nt s in MRI produces large osci llating Lorentz forces which in turn produce co il vibrations aga in st the magnet cryostat with an assoc iated rat-tal noise. T hi s

346 INDIAN J. BIOCHEM . BIOPI-IYS., VOL. 36. OCTOBER 19<)<)

repetitive sound is enhanced by higher gradient duty cyc les and sharper pul se transitions. Acoust ic noi se is thus likely to increase with decreases in sec ti on thicknesses, decreased fi e lds of vi ew, repetiti on times and echo times . The noise leve ls in MR scanners are in the ran ge of 65-95 decibels (d B), which is considered to be within the recommended safety guidelines. However, there ha ve been reports that acoustic noi se generated during MRI inves tigation causes patient annoyance, interference wit h oral communication, and reversibl e hearing loss in patients

I d . d . ?6 '7 S If ' w 10 I not wear ear protect lon- '- . tue yo ' patients undergo ing MRI without car plugs showed temporary hearing loss in 43 % of the subjects26 Strategy for reducing sound leve ls during MRI is bei ng worked out. One method is to use an "antinoi se" or destructive interference tec hnique that not onl y effectively reduces noise but also permits better pat ient communic ati o n 2~ . A recent stud y found no signifi cant degradation of MR Image quality with the use of antinoise method28

.

MRI during pregnancy Although MRI is not believed to be hazard ous to

the fetu s, on ly a few investi ga ti ons exa mining the teratogenic potential of thi s imagin g modality are reported. Most of the earli er studies conducted to determine poss ible unwanted bioe ffects during

I d . I 09 10 R . ,. pregnancy s lowe negati ve resu ts- ·· . ecent stue les on mi ce exposed during mid ges tati on did not show gross embryotox ic effects but there was a reduction in

' 0 crown-to-rump length .' . A va ri ety of mec hanisms ex ist th at could produce

deleteri ous bioeflects with respect to the deve loping fetus and the use of electromagnetic fi elds during MRI' U2. Patients who are pregnant or who suspected to be pregnant are to be assessed for the risks versus the benefit s of the MR examination . Care shou ld be exercised during the first trimester to avoid medioco lega l implicati ons relati ve to spontaneous abortions because there is a hi gh spontaneous abortion rate in general population during the first trimester of pregnancy (i .e.,>30%) . Hence, a cauti ous approach is recommended for the use of MRI is pregnant patients .

General considerations

Approximately 5% to 10% of patients undergoing MRI have c laustrophobia and a va riety of other psychological sensations including anxiety and panic disorders. These are due to the narrow interior (tunnel) of the scanner, the duration of the

examination, grad ient induced noise, c tc,Y17 Mos t of

these psychological responses , are u s u ~t1l y transient. Relaxation methods and detailed briefing of the MRI procedure are found to red uce the anxiety leve l to certain extent. In recent times, M R system arch itectures offer a more open des ign that reduce c laustroph obi a and the frequency or psyc hological problems associated wit h MRI procedures .

Overall, numerous studi es perrormcd spec ifi call y to study the potential bioe ffe cts or MR I have revea led negative results supporting the widely he ld vicw th at there is no significant health risks assoc iated with the use of MRT modalit/. Howeve r, as disc ussed earlier 111 thi s artic le, patien ts with inte rnal Glrdiac pacemakers , implantable ca rdiac defibrillators, coc hler implan ts, neurost i mulators, bone-growth sti mulato rs, implantable e lectron ic drug infusion pumps, and other simi lar dev ices that co uld ad verse ly be affected by the magneti c fi eld should not be exami ned by thi s imaging t ec hnique ~7.3~· ·"'. Similarl y, MRI is contraindicated 1'0 1' pati ents who ha ve ce rt ain ferromagnetic implants, mate rial s, or roreign bodi es, primaril y because of the poss ibility 01' movement or dislodgement of th ese objects('·x. Pati ent s \V ith non fe rromagnetic aneurys m and hemos t,ll ic eli ps can be imaged. Similarly , one should fir st make sure th at carotid artery vascular clamps, dental dev ices an d material s, heart valves, intravascu lar co il s, rilt ers and stents , occu lar implants, orthopedi c imp lant s, otolog ic implants , etc. are to be tested for fL:: rromagneti sm before MR in ves ti gation . These are cas il y achi eved by written screening and through metal detec tor tes ts.

Conclusion

The possibl e health hazards rclated to IfR f inves ti gationa l procedure under dillerent catego ri es presented above have been dul y taken into consideration by the equipment manu Llct urers and necessary prevent ive measures arc usual! y incorporated into the scanner. The personne l connected with MRf si te are to be well-tra ined in emergency procedures. The awareness 0" the numerous areas of concern regardin g IfRI environments both for patient s undergo in g fVIR examination and for MR health practiti oners is hi ghl y desirabl e. An understandin g of these potential interactions allows a more knowledgeable dec ision to be made because fVIR I is now a frequentl y recommended early diagnostic modality 1'0 1' man y so rt tissue pathologies .

JAGA NNATI-IAN : MAGNETIC RESONANCE IMAGING: BIOEFFECTS AND SArETY CO NCERNS 347

References Raghunath an P & Jagannathan NR( 1996) Cll rr Sci 70, 698-708

2 Shellock F G & Kanal E( 1997) in Magn e/ic Resonance Imaging of the Body 3rd edn (I-liggins C B, I-Iricak H & I-I e lms C A eds), pp 175-204 and references therein, Lippincott - Raven , Philadelphi a

3 Tenforde T S(1986) Bioelectromagnetics 7, 341-346 4 Shellock F G, Schaefer D .J & Gordon CJ( 1986) Magn Reson

Med 3,644-647 5 Shellock F G, Schaefer D J & Crues JV ( 1989) Magn Reson

Med 11,371-375 6 Shellock F G & Crues J V ( 1988) MRI Decisions 2, 25-30 7 Shellock F G (1989). Magn Reson Q 5, 243-261 8 Kanal E, Taiagala L & Shallock F G ( 1990). Radiology 176,

543-606 9 Tenforde T S , Gaffey CT, Moyer B R & Budinger T F( 1983)

Bioelectromagnetics 4, 1-4 10 Reding ton R W, Dumoulin C L & Schenck J L (1988) [n

Book of Abstracts Society of Magnetic Resonance Medicine I , 20

II Reilly J P (1989) Med Bioi Eng Compllt27, 101-112 12 Bernhardt J (1979) Radiat Environ Biophys 16, 309-323 13 BudingerT F (1981) J Complll Assist TOll1ogr 5,800-811 14 Budinger T F, Cullander C & Bordow R (1984). In : Book of

Abstracts Society of Magnetic Resonance Medicine 15 Modan B (1988) Am J Indian Med 13, 625-627 16 Brown H D & Chattopadhyay S K (1988, Cancer Biochem

Biophys 9, 295-342 17 Pool R ( 1990) . Science 249, 1378- 138 1 18 Gordon C J ( 1987) Bioelectromaglletics 8, 111-118 19 Beers J ( 1989) Magn Reson Imaging 7, 309-33 I 20 Coulter J S & Osbourne S L ( 1936) Arch Phys Th er 17,679-

687 2 1 Bottomley P A & Edelstein W A ( 198 1). Med Phys 8, 510-

512

22 Bottomley P A, Redington R W & Edelste in W A ( 1985 ) Magn Reson Med 2. 336-349

23 Shellock F G, Schaefer D J & Crues J V (19))4) /Jr.l Rod 62 . 904-909

24 Shellock F G, Rothman B & Sarti D ( 1990) Am .I Roentgenol 154, 1229-1232

25 Shellock F G & Crues J V ( 1996) Radiology 167. g09- I I 26 Brummett R E, Talbot J M & Charuhes P ( 19XX) Radiolo?."

169,539-540

27 Shellock F G & Kanal E (1991) J Magn Re.l'on III/aging I , 97-101

28 Goldman A M , Grossman W F & Friedlander P C ( 1989 ) Radiology 173, 549-554

29 Geard C R , Osmak R S, I-I a ll E J. Simon 1-1 E, Maudsley AA & Hilal SK (1984) Radiologv 152 , 199-202

30 Heinrichs W L, Fong P, Flannery M , I-I enricks S C. Crooks L E, Spindle A & Pedersen R D ( 1988) Magn Reson III/aging 6, 305-313

31 Barnothy M F ( 1969), Biological ellects oj' II/agnetic Feltls Vols I and 2, Plenum Press, New York

32 A dey W R (1981) Physioi Rei' 61 , 435-514

33 Flaherty J A & Hoskinson K ( 19X9) Ne lv Engl .I Med 320, 467-468

34 Fishbain D A, Goldberg M . Labbe E, Zacher D, Steel Rosomoff R & Rosomaff 1-1 (1988) Alii .I Psrhiarrv 145 , 1038-1039

35 Quirk M E , Latendre A J , C iottone R A & Lingl ey JF ( 1989) Radiology 170, 463-466

36 Quirk M E, Latendre A J , Ci ottone R A & Lingley J F ( 19R9) Radiology 173,759-762

37 Hircak H & Amparo E G (1984) Radiologl' 152 , X IlJ

38 Hayes D L, Ho lmes D R & Grag J E ( l lJX 7) .I Alii Coli Cardio110,782-786

39 Gangarosa R E, Minni s J E, Nobbe J , Prasc han D & Gcnherg R W ( 1987) Magn Resonllllaging 5 287-21)2 .