- 101 -

JKSMRM 9:101-108(2005)1Department of Diagnostic Radiology and Research Institute of Radiological Science and 2Pulmonary Division, Department ofInternal Medicine, Yonsei University, College of Medicine, Yong Dong Severance HospitalThis paper was supported by the Brain Korea 21 Project for Medical Science, Yonsei UniversityReceived; October 20, 2005, accepted; November 15, 2005Address reprint requests to :Tae-Sub Chung, M.D., Department of Diagnostic Radiology and Research Institute of Radiological

Science, Brain Korea 21 Project for Medical Science, Yonsei University, College of Medicine, YongDong Severance Hospital, 146-92 Dogok-dong, Kangnam-gu, Seoul 135-270, South Korea.Tel. 82-2-2019-3514 Fax. 82-2-3462-5472 E-mail: tschung@yumc.yonsei.ac.kr

Evaluation between 3.0 T vs 1.5 T MRI in Detection ofBrain Metastasis using Double Dose Gd-DTPA

Woo-Suk Chung, M.D.1, Tae-Sub Chung, M.D.1, Hyung Jung Kim, M.D.2, Chul Min Ahn,M.D.2, Jae Hoon Lee, M.D.1, Jin Hur, M.D.1, Arthur Eung-Hyuck Cho, M.D.1

Purpose : Early detection of small brain metastases is important. The purpose of thisstudy was to compare the detectability of brain metastases according to the sizebetween 1.5 T and 3.0 T MRI.Materials and Methods : We reviewed 162 patients with primary lung cancer whowere examined for TNM staging. After administration of double dose of Gd-DTPA,MR imaging was performed with SPGR by 3.0 T MRI and then with T1 SE sequenceby 1.5 T MRI. In each patient, three readers performed qualitative assessment.Sensitivity, positive predictive value, and diagnostic accuracy were calculated in 3.0T and 1.5 T MRI according to size. Using the signal intensity (SI) measurementsbetween the metastatic nodules and adjacent tissue, nodule-to-adjacent tissue SI ratiowas calculated. Results : Thirty-one of 162 patients had apparent metastatic nodules in the brain ateither 1.5 T or 3.0 T MR imaging. 143 nodules were detected in 3.0 T MRI, whereas137 nodules were detected at 1.5 T MRI. Six nodules, only detected in 3.0 T MRI,were smaller than 3.0 mm in dimension. Sensitivity, positive predictive value, anddiagnostic accuracy in 3.0 T MRI were 100 %, 100 %, and 100 % respectively, andin 1.5 T MRI were 95.8 %, 88.3 %, and 85.1 % respectively. SI ratio was significantlyhigher in the 3.0 T MRI than 1.5 T MRI (p=0.025). Conclusion : True positive rate of 3.0 T MRI with Gd-DTPA was superior to 1.5 TMRI with Gd-DTPA in detection of metastatic nodules smaller than 3.0 mm.

Index words :Magnetic resonance (MR), high-field-strength imagingBrain neoplasm, metastasesContrast medium

Introduction

Brain metastases present a poor prognosis suggestinga shortened survival time. Early diagnosis of braininvolvement and determination of the number ofmetastases are important not only for quality of life butalso for cost effectiveness (1-3). The decision regardinga conservative versus a surgical approach depends onthe number of brain metastases detected by radiologicmeans (1-3).

A contrast-enhanced MRI has become the method ofchoice for visualization of brain metastases (1). High-dose gadolinium-enhanced MR examinations may haveadvantages over 0.1 mmol/kg examinations in detectingearly and/or small metastases (3-7). A delayed studycan also increase the contrast (2). Comparing to the 1.5T MRI with an axial T1-weighted spin echo sequence(SE), the higher field strength MR systems combinedwith a sequence of SPGR (spoiled gradient recalledacquisition in the steady state) and administration of ahigh dose of Gd-DTPA (gadopentetate dimeglumine)may also have advantages in detecting smallmetastases, although this has not yet been verified byclinical data. The purpose of this study is to comparethe detectability of brain metastases classifiedaccording to the size of nodules, between thin sliceSPGR of 3.0 T and conventional thick slice SE of 1.5 TMRI with the administration of a double dose of Gd-DTPA.

Materials and Methods

PatientsFrom December 2002 to February 2004, a total of 162

consecutive patients with primary lung cancerparticipated in our study. The institutional reviewboard approved our study, and informed consent wasobtained from all patients regarding the potential risksof both the double dose of contrast medium and twoassessments by MRI scanning on a 3.0 T and 1.5 Tmachine. After the study, 31 patients of this populationwere diagnosed as having brain metastases. Meanpatient age was 61.2 (range, 43 to 80 years).

ProtocolOn the 3.0 T MR scanner (GE Signa VH/i; GE

medical system, Milwaukee, USA), the images wereacquired using a standard head coil and an activelyshielded gradient system with a maximum gradientstrength of 43 mT/m. On the 1.5 T MR scanner(Magnetom Vision; Siemens Medical Systems,Erlangen, Germany), a standard head coil and amaximum gradient strength of 25 mT/m were used.

All patients were examined after administrating acontrast agent with a double dose of Gd-DTPA (0.2mmol/kg). First, examinations were performed on a 3.0T MR scanner and then were subsequently performedon a 1.5 T MR scanner without additional contrastinjection. The scan interval between the 3.0 T and 1.5T MR examination was less than 20 minutes. Thecontrast agent used was Gd-DTPA (Magnevist;Schering AG, Berlin, Germany). In all patients, thedouble dose of Gd-DTPA at 0.2 mmol/kg wasadministered intravenously as a bolus and scanningcommenced immediately.

MR imaging included the following sequences onboth scanners: At 3.0 T MRI, an axial 3D SPGR, whichis usually used at present, was used (TR/TE/TI =5.7/1.44/400 milliseconds; flip angle 20°; 2 mm slicethickness; FOV of 220 mm; matrix size of 512×512ZIP; spatial resolution of 0.43×0.43×2 mm; twoacquisition) with a scan time of 3 minutes 30 seconds.At 1.5 T MRI, an axial T1-weighted spin echo sequencewas used (TR/RE = 600/14 milliseconds; flip angle 90°;5 mm slice thickness; FOV of 210 mm; matrix size of174×256; spatial resolution of 1.2×0.82×5 mm; twoacquisition) with a scan time of 5 minutes 30 seconds.

Three radiologists performed randomized,independent blinded review. The three readers weremerely informed that all of the patients had lungmalignancies. The postcontrast MR examinations of thebrain on all of the patients were evaluated. The readersdid not have access to other image sets within eachstudy.

The presence, size, and number of metastatic noduleswere assessed. The postcontrast images were dividedinto the following two groups: 3.0 T MRI with a doubledose of Gd-DTPA and 1.5 T MRI with a double doseof Gd-DTPA. The readers were asked to document thenumber of nodules. Each nodule in each study wasnumbered and classified according to its largestdiameter measurement: ≤3 mm, 3 mm to 5 mm, or>5 mm.

Woo-Suk Chung et al

- 102 -

Subsequently, if there was any debate about noduleclassification, a final interpretation using imagingstudies was done until consensus among the readerswas accomplished. If disagreements persisted, 3-6months of follow-up MR scans were obtained forfurther validation. For the patients who had follow-upMR scans, these scans were considered positive formetastases if there was a response to the treatment orif there was a growth of nodules identified during thefollow-up period.

In addition, the sensitivity, specificity, positivepredictive value, negative predictive value, anddiagnostic accuracy of the 3.0 T and 1.5 T MRI werecalculated using the results from the final interpretationaccording to size.

A quantitative image assessment was performed next.The same lesions on scans of the two different MRIsystems were determined by comparing peripheralstructures. Signal intensities of adjacent tissue andnodule were assessed by region of interest (ROI)measurements placed identically on both series ofimages using the same sized circular ROI from thesoftware available on both scanners. Nodule-to-adjacent tissue contrast (SI ratio) is defined by

SI ratio = ×100

Stissue1 taken as the signal intensity of an ROI assessedover a nodule and Stissue2 taken as the signal intensityassessed over the contralateral white matter. The signalintensity of a nodule was measured within anenhanced area. In inhomogeneously enhanced nodules,the area of maximum uptake was chosen formeasurement.

Statistical AnalysisThe paired Wilcoxon’s signed ranks test was used to

compare qualitative scores, and the matched-pair-t testwas used to compare nodule SI ratios between the 3.0T and 1.5 T MRI with a double dose of Gd-DTPA. Forall tests, significance was set at p < 0.05, and SPSSsoftware (SPSS Inc. Chicago, US) was used forstatistical analyses.

Results

In the final review made by consensus of three

radiologists (Table 1), a total of 143 metastatic noduleswere detected by the 3.0 T MRI with a double dose ofGd-DTPA. Of these nodules, 49 nodules were ≤ 3 mmin diameter, 37 nodules were 3 mm to 5 mm, and 57nodules were >5 mm. A total of 137 metastaticnodules were detected by the 1.5 T MRI with a doubledose of Gd-DTPA in the final interpretation. Of thesenodules, 43 nodules were ≤ 3 mm in diameter. The3.0 T MRI with a double dose of Gd-DTPA wassignificantly more effective at detecting nodules smallerthan 3 mm (p = 0.014) than the 1.5 T MRI. However,we found no significant difference in the detection ofnodules larger than 3 mm (p > 0.05) (Table 1).

In 6 of the 31 cases, six nodules, which were onlydetected by the 3.0 T MRI with a double dose of Gd-DTPA, and missed by the 1.5 T MRI with a doubledose of Gd-DTPA, were unanimously agreed upon byall participants in the final interpretation. In three ofthe six nodules, a growing nodule was detected duringa follow-up MR scan which confirmed the presence ofmetastases (Fig. 1). Three of the 6 nodules missed wereunable to be detected using the 1.5 T MRI with adouble dose of Gd-DTPA due to artifact (Fig. 2). Finallythe 3.0 T MRI with a double dose of Gd-DTPA was

Stissue1 - Stissue2

Stissue2

Evaluation between 3.0 T vs 1.5 T MRI in Detection of Brain Metastasis using Double Dose Gd-DTPA

- 103 -

Table 1. Total Number of Brain Metastases Classified by Sizeand Detected at the Final Interpretation

Size3.0 T MRI with 1.5 T MRI with

p valueGd-DTPA Gd-DPTA

≤ 3 mm 49 43 0.0143 mm<-≤5 mm 37 37 > 0.05

5 mm< 57 57 > 0.05Total 143 137 0.014

Note : Data are the number of nodules

Table 2. The Values of Currently Used Statistical Measures forthe 3.0 T and 1.5 T MRIs with Gd-DTPA According to NoduleSize

Metastatic nodules

Total ≤ 3 mm

Present Absent Present Absent

3.0 T MRI with Positive 143 00* 49 0*Gd-DTPA Negatiive 000 00* 00 0*1.5 T MRI with Positive 137 18* 43 18*Gd-DTPA Negative 006 00* 06 0*

* Pseudolesions that were detected as nodules by the 1.5 T MRIwere detected as vascular structures by the 3.0 T MRI.

useful for confirming the presence of these nodules.Nodules observed by the 3.0 T MRI with a double

dose of Gd-DTPA were brighter or better delineationthan by the 1.5 T MRI with a double dose of Gd-DTPA.Out of 137 nodules, 65 were brighter (Fig. 3) and out of137 nodules, 63 were better delineated (Fig. 4). Forty-five of 137 nodules were satisfactory in both conditions.

By consensus, the final review stated, nodulesdetected by the 3.0 T and 1.5 T MRI were considered

positive for metastases (Table 2). Additionally, therewere 18 pseudolesions which were detected as nodulesby the 1.5 T MRI, but as vascular structures by the 3.0T MRI (Fig. 5). The sensitivity, positive predictivevalues, and diagnostic accuracy of the 1.5 T MRI witha double dose of Gd-DTPA were 95.8%, 88.3%, and85.1% respectively. The sensitivity, positive predictivevalue, and diagnostic accuracy of the 3.0 T MRI witha double dose of Gd-DTPA were 100%, 100%, and

Woo-Suk Chung et al

- 104 -

a bFig. 2. A 73-year-old male with multiple brain metastases. Metastatic nodule (a) by a 3.0 T MR image with a doubledose of Gd-DTPA show a metastatic nodule (arrow). But (b) the 1.5 T MR image with a double dose of Gd-DTPAcannot show the metastatic nodule by artifact.

a b cFig. 1. A 65-year-old male with single brain metastasis. (a) The 3.0 T MR image with a double dose of Gd-DTPAshows a small metastatic nodule (arrow). (b) The 1.5 T MR image with a double dose of Gd-DTPA cannot showthe nodule. In (c), a growing nodule was detected during the follow-up MRI with a double dose of Gd-DTPA whichconfirmed the presence of metastasis (arrow).

100% respectively. For metastatic nodules smaller than3 mm, the sensitivity, positive predictive value, anddiagnostic accuracy of the 1.5 T MRI with a doubledose of Gd-DTPA were 87.8%, 70.5%, and 64.2%respectively. The sensitivity, positive predictive value,and diagnostic accuracy of 3.0 T MRI with a doubledose of Gd-DTPA were 100%, 100%, and 100%respectively (Table 3).

For quantitative image assessment, the SI ratio in thepost contrast sequences (Table 4) was significantlyhigher in the 3.0 T MRI with a double dose of Gd-DTPA than in the 1.5 T MRI with a double dose of Gd-

DTPA (p = 0.025).

Discussion

Detection of metastatic nodules is dependent on boththeir size and contrast ratio (2). As methods to increasethe contrast, a higher dose of Gd-DTPA, a higher fieldstrength, and delayed study can all be used to aiddetection (2-5, 8-17). Nodules larger than 10 mm areeasily detected because vasogenic edema is customarilyassociated with larger metastases (2). So, when a higherdose of Gd-DTPA is used, the detection rate of largernodules is not influenced. However, a higher dose ishelpful for detecting small nodules because it increasesnodule enhancement, yet this method has thedisadvantages of increasing the false positive rate andpromoting side effects (2-5, 8-14). A delayed studycan also increase the contrast (2). It has beenrecommended that image acquisition be delayed from5 to 35 minutes after the administration of contrastmaterial at a dose of 0.1 mmol/kg to ensure optimaldetection (2, 15).

Three of the 6 nodules missed were unable to bedetected using the 1.5 T MRI with a double dose of Gd-DTPA due to partial volume artifact. Eighteen lesionswere detected in the 1.5 T MRI with a double dose ofGd-DTPA, which proved to be vascular structures in3.0 T MRI with a double dose of Gd- DTPA. The truenature of these lesions was revealed because of thegreater morphologic detail visualized by the high field

Evaluation between 3.0 T vs 1.5 T MRI in Detection of Brain Metastasis using Double Dose Gd-DTPA

- 105 -

a b

Fig. 3. A 65-year-old male withmultiple brain metastases. (a) The3.0 T MR image with a doubledose of Gd-DTPA shows ametastatic nodule with sphericalenhancement (arrow). (b) The 1.5T MR image with a double doseof Gd-DTPA shows a metastaticnodule with spherical enhance-ment (arrow). Nodules observedby the 3.0 T MRI with a doubledose of Gd-DTPA were brighterthan when observed by the 1.5 TMRI with a double dose of Gd-DTPA.

Table 3. Estimation of the Values of Currently Used StatisticalMeasures for the 3.0 T and 1.5 T MRI with Gd-DTPA Accordingto Nodule Size

Metastatic nodules

Total ≤ 3 mm

3.0 T 1.5T 3.0 T 1.5T

Sensitivity 100 95.8 100 87.8Specificity - 0 - 0Positive predictive value 100 88.3 100 70.5Negative predictive value - 0 - 0Diagnostic accuracy 100 85.1 100 64.2

Table 4. Results of a quantitative assessment of nodules thatwere detected by the 3.0 T and 1.5 T MRIs with Gd-DTPA(n=137)

3.0 T 1.5 T P value

Nodules-to-adjacent tissue: SI ratio 94.59 63.86 0.025

strength MR image, thus allowing differentiationbetween true enhancing lesions and sulcal vessels.

In our study, each nodules was classified accordingto its largest diameter as being ≤ 3 mm, 3 mm to 5mm, and >5 mm because the slice thickness of the 3.0T MRI was 2.0 mm and 5 mm for the 1.5 T MRI.Therefore this study classified small metastatic nodulesas being smaller than 3 mm.

Although our study supports the use of a higher fieldstrength MRI with a double dose of Gd-DTPA forincreased metastatic nodule detection and for improvednodule enhancement and delineation, the results

should be interpretated with caution. The reason forthis is that for patients with two or more brainmetastases, additional metastases found with the 3.0 TMRI seem to be of limited clinical importance. Thepresence of two or more small nodules generally willnot change the way the patient is managed. Therefore,it is of utmost importance to identify the differencebetween none, one, and more than one metastaticnodule. Patients with a single metastatic nodule locatedin a respectable region can be treated surgically, andthe tumor staged as M1, not M0. However, patientswith two or more metastatic nodules are usually

Woo-Suk Chung et al

- 106 -

a b

Fig. 5. A 67-year-old female withmultiple brain metastases. (a) The3.0 T MRI with a double dose ofGd-DTPA shows vascularstructure (arrow). (b) The 1.5 TMRI with a double dose of Gd-DTPA shows a nodular lesion(arrow). A pseudolesion that wasdetected as a nodule by the 1.5 TMRI with a double dose of Gd-DTPA but as a vascular structureby the 3.0 T MRI with a doubledose of Gd-DTPA is showed inthis figure.

a b

Fig. 4. A 64-year-old male withmultiple brain metastases. (a) The3.0 T MRI with a double dose ofGd-DTPA shows a metastaticnodule with ring enhancement(arrow). (b) The 1.5 T MRI withGd-DTPA show a metastaticnodule with ring enhancement(arrow). Nodules observed by the3.0 T MRI with a double dose ofGd-DTPA were better delineatedthan when observed by the 1.5 TMRI with a double dose of Gd-DTPA.

treated with radiation therapy and/or systemicchemotherapy (3, 18-22). On the other hand, the useof a higher strength field MRI was found to be helpfulin confirming the appearance of an equivocalmetastatic nodule. The 3.0 T MRI with a double doseof Gd-DTPA was also useful for the detection ofadditional metastases in patients with a known lesiondetected by the 1.5 T MRI with a double dose of Gd-DTPA. Because this also has influence on thesensitivity, positive predictive value, and diagnosticaccuracy the 3.0 T MRI with a double dose of Gd-DTPA was found to have better results than the 1.5 TMRI with a double dose of Gd-DTPA in our study.Therefore we recommend the use of the 3.0 T MRIwith a double dose of Gd-DTPA in only threecircumstances: when the findings by the 1.5 T MRIwith Gd-DTPA are equivocal, when one potentiallysurgically respectable nodule is identified, or fordetecting early and/or small metastases

There were two limitations in this study. The firstlimitation was a difference in protocol sequence. TheT1 weighted spin-echo protocol was used for the 1.5 TMRI and SPGR technique protocol for the 3.0 T MRI.Detectability of metastatic nodules is more effectivewith 3.0 T MRI with T1 SE than with a 1.5 T MRI withSE (2). SPGR with thin slice thickness is superior to theT1 spin echo sequence with thick slice thickness dueto a partial volume effect. Additionally, SE is notoptimum for a 3.0 T due to longer T1- and shorter T2-relaxation times of water protons, which decrease thecontrast ratio in the 3.0 T images.17 The purpose ofthis study was to detect early small brain metastases,so using the SPGR sequence with a 3.0 T MRI, whichis usually used at present, has advantages for detectingearly small brain metastases. The second limitation wasthat the scan interval between the 1.5 T and 3.0 T MRIwas less than 20 minutes. However, a delayed studyincreases contrast as mentioned above. The currentstudy found that the SI ratio was significantly higher inthe 3.0T images than in the 1.5 T images. Consideringthe delayed study by the 1.5 T MRI, this limitationemphasizes the better detection rate of the 3.0 T MRIwith a double dose of Gd-DTPA. And due to theseconfounding variables, blinded reviews wereperformed by three readers.

The detectability of metastatic nodules smaller than3mm was better using a 3.0 T MRI with SPGR than a

1.5 T MRI with T1 SE. Therefore we recommend a 3.0T MRI with SPGR and a double dose of Gd-DTPA fordetecting early and/or small metastatic nodules,furthermore influencing treatment.

References

1.Lassman AB, DeAngelis LM. Brain metastases. Neurol Clin.2003; 21:1-23, vii, Review

2.Ba-Ssalamah A, Nobauer-Huhmann IM, Pinker K, et al. Effectof contrast dose and field strength in the magnetic resonancedetection of brain metastases. Invest Radiol 2003; 38: 415-422

3.Sze G, Johnson C, Kawamura Y, et al. Comparison of single-and triple-dose contrast material in the MR screening of brainmetastases. AJNR Am J Neuroradiol 1998; 19:821-828

4.Runge VM, Kirsch JE, Burke VJ, et al. High-dose Gadoteridolin MR imaging of intracranial neoplasm. J Magn ResonImaging 1992; 2:9-18

5.Yuh WT, Engelken JD, Muhonen MG, Mayr NA, Fisher DJ,Ehrhardt JC. Experience with high-dose gadolinium MRimaging in the evaluation of brain metastases. AJNR Am JNeuroradiol 1992; 13:335-345

6.Brekenfeld C, Foert E, Hundt W, Kenn W, Lodeann KP, GehlHB. Enhancement of cerebral diseases: How much contrastagent is enough? Comparison of 0.1, 0.2, and 0.3 mmol/kgGadoteridol at 0.2 T with 0.1 mmol/kg Gadoteridol at 1.5 T.Invest Radiol 2001; 36:266-275

7.Yuh WT, Tali ET, Nguyen HD, Simonson TM, Mayr NA,Fisher DJ. The effect of contrast dose, imaging time, andlesion size in the MR detection of intracerebral metastasis.AJNR Am J Neuroradiol 1998; 16:373-380

8.Van Dijk P, Sijens PE, Schmitz PI, Oudkerk M. Gd-enhancedMR imaging of brain metastases: contrast as a function ofdose and lesion size. Magn Reson Imaging 1997; 15:535-541

9.Yuh WT, Fisher DJ, Engelken JD, et al. MR evaluation ofCNS tumors: dose comparison study with gadopentetatedimeglumine and gadoteridol. Radiology 1991; 180:485-491

10.Healy ME, Hasselink JR, Press GA, Middleton MS. Increaseddetection of intracranial metastases with intravenous Gd-DTPA. Radiology 1987; 165:619-624

11.Akeson P, Vikhoff B, Stahlberg F, Holtas S. Brain lesioncontrast in MR imaging: dependence of field strength andconcentration of gadodiamide injection in patients andphantoms. Acta Radiol 1977; 38:14-18

12.Runge VM. A review of contrast media research in 1999 -2000. Invest Radiol 2001; 36:123-130

13.Yuh WT, Parker JR Carvlin MJ. Indication related dosing formagnetic resonance contrast media. Eur Radiol 1977; 7(suppl5):S269-S275

14.Rinck PA, Muller RN. Field strength and dose dependence ofcontrast enhancement by Gadolinium-based MR contrastagents. Eur Radiol 1999; 9:998-1004

15.Schorner W, Laniado M, Niendorf HP, Schubert C, Felix R.Time-dependent changes in image contrast in brain tumorsafter gadolinium-DTPA. AJNR Am J Neuroradiol 1986;

Evaluation between 3.0 T vs 1.5 T MRI in Detection of Brain Metastasis using Double Dose Gd-DTPA

- 107 -

7:1013-102016.Chang KH, Ra DG, Han MH, Cha SH, Kim HD, Han MC.

Contrast enhancement of brain tumors at different MR fieldstrengths: comparison of 0.5 T and 2.0 T. AJNR Am JNeuroradiol 1994; 15:1413-1419

17.Nobauer-Huhmann IM, Ba-Ssalamah A, Mlynarik V, et al.Magnetic resonance imaging contrast enhancement of braintumors at 3 tesla versus 1.5 tesla. Invest Radiol 2002; 37:114-119

18.Galicich JH, Sundaresan N, Thaler HT. Surgical treatment ofsingle brain metastasis: evaluation of results by computerizedtomography scanning. J Neurosurg 1980; 53:63-67

19.Runge VM, Carollo BR, Wolf CR, Nelson KL, Gelblum DY.

Gd-DTPA: A review of clinical indications in central nervoussystem magnetic resonance imaging. Radiographics 1989;9:929-958

20.Earnest F 4th, Ryu JH, Miller GM, et al. Suspected non-smallcell lung cancer: Incidence of occult brain and skeletalmetastases and effectiveness of imaging for detection-Pilotstudy. Radiology 1999; 211:137-145

21.Koutras AK, Marangos M, Kourelis T, et al. Surgicalmanagement of cerebral metastases from non-small cell lungcancer. Tumori 2003; 89:292-297

22.Soffietti R, Ruda R, Mutani R. Management of brainmetastases. J Neurol 2002; 249:1357-1369. Review

Woo-Suk Chung et al

- 108 -

통신저자 : 정태섭, 서울특별시 강남구 도곡동 146-92 연세대학교 영동세브란스병원 영상의학과Tel. 82-2019-3514 Fax. 82-3462-5472 E-mail: tschung@yumc.yonsei.ac.kr

뇌전이종양의발견에있어서Doble dose Gd-DTPA를이용한3 T MRI와1.5 T MRI간의비교연구

1연세대학교 영동세브란스병원 영상의학과2연세대학교 영동세브란스병원 호흡기내과

정우석1·정태섭1·김형중2·안철민2·이재훈1·허 진1·조응혁1

목적: 작은 뇌전이 종양의 조기 발견은 중요하다. 이 연구의 목적은 1.5 T MRI와 3.0 T MRI 간의 크기에 따른 뇌

전이 종양의 발견율을 비교하는 것이다.

대상 및 방법: 폐암으로 진단 받은 162명의 환자를 대상으로 TNM 병기를 위해 뇌 MRI를 시행하였다. Gd-DTPA

를 2배 용량으로 투여 후, 3.0 T MRI에서 훼손경사회복획득으로 촬영하였으며 그 후 1.5 T MRI에서 T1 스핀 에

코로 촬영하였다. 3명의 방사선과 전문의가 합의하여 MRI를 판독하였으며 정성 평가를 시행하였다. 3.0 T와 1.5 T

MRI에서 크기에 따라 민감도, 양성 예측률, 정확도를 평가하였다. 신호 강도를 사용하여 전이 종양과 인접 조직간

의 신호강도 비를 계산하였다.

결과: 162명의 환자 중 31명에서 1.5 T 또는 3.0 T MR에 뇌전이 종양이 발견되었다. 3.0 T MRI에서 143개의 종

양이 발견되었으나 1.5 T MRI에서 137개의 종양이 발견되었다. 6개의 종양이 3.0 T MRI에서만 발견되었으며 크

기는 모두 3 mm 미만이었다. 3.0 T MRI의 민감도, 양성 예측률, 정확도는 각각 100%, 100%, 100%이며, 1.5 T

MRI에서는 각각 95.8%, 88.3%, 85.1% 이다. SI ratio는 1.5 T MRI보다 3.0 T MRI에서 유의하게 높았다

(p=0.025).

결론: Double dose Gd-DTPA를 이용한 3.0 T MRI는 3 mm미만의 뇌전이 종양을 발견하는데 있어서 1.5 T MRI

보다 우수하다.

대한자기공명의과학회지 9:101-108(2005)