목 적：고자장(3.0T) MRI에서 교정 후 잔존하는 강자성체 인공물에 대해 SEMAC 기법의 단계별 적용을 통하여 T1, T2 검사 시퀀스의 축상면 인공물의 장･단축 길이 감소 정도와 신호대 잡음비 측정을 통하여 인공물 감소를 위한 최적의 단계를 알아보고 임상에서 추가적인 검사 방법으로 적용하고자 한다.
대상 및 방법：3.0T MRI (MAGNETOM Skyra, Siemens, Munich, Germany)를 사용하여 자체 제작된 손목(Wrist and hand) 팬텀 속에 치료용 보루스와 치과용 stainless steel wire (18 × 25 mm)를 삽입하여 고정하였다. 고신호 강도를 구현하기 위해서 두･부 전용 코일(64 channel)을 사용하여 검사를 진행하였으며, 연구에 사용한 펄스 시퀀스는 T1 TSE, T2 TSE에 SEMAC 기법을 적용하였고, 추가적인 위상 부호화 단계(phase encoding steps, PES)를 정성적(6-15), 정량적(6-10)까지 변화시켜 10회 반복 측정하여 실험하였다. 정량적 평가는 영상 왜곡이 가장 심하게 일어난 부위(영상 18번) 에서 좌･우측으로 나누고 장･단축의 길이를 계측하였고, 인공물 영향이 없는 3곳을 좌･우측 각각 지정하여 신호대 잡음비 (signal to noise ratio, SNR)를 계측하였다. 정성적 평가는 이미지의 질을 내･외부 평가자 각각 3명이 영상 평가 기준에 맞춰 5점 척도화하여 평가하였다.
결 과：T2 SEMAC의 인공물에 대한 정량적 분석 결과는 PES가 6→7, 7→8, 8→9, 9→10 변화할 때 RT 장축 길이는 0.11%, 0.02%, 0.10%, 0.02%로 감소, 단축 길이는 0.19%, 0.04%, 0.22%, 0.07%로 감소하였다. LT 장축 길이: 0.12%, 0.02%, 0.10%, 0.06%로 감소, 단축 길이: 0.20%, 0.09%, 0.18%, 0.3%로 감소하였다. T1 영상의 정량적 분석의 경우는 RT의 장축 길이: 0.17%, 0.01%, 0.11%, 0.01%로 감소, 단축 길이는 0.14% 0.01%, 0.11%, 0.02%로 감소하였다. LT의 장축 길이: 0.20%, 0.01%, 0.09%, 0.01% 감소, 단축 길이: 0.13%, 0.03%, 0.11%, 0.01%로 감소되는 결과를 나타내었다 (p<0.01). T2(RT)의 신호대 잡음비 측정 결과 PES가 6-10까지 증가할 때 101.92, 105.25, 105.44, 104.44, 103.47, T2(LT): 95.30, 98.98, 97.22, 96.61, 95.74, T1(RT): 177.24, 175.50, 296.06, 299.88, 313.71이고, T1(LT): 159.67, 158.79, 246.75, 226.75, 259.67로 나타났다. 정성적 분석의 경우 T2 영상에서 5점 척도를 기준으로 SEMAC PES가 6~15일 경우 1.50, 2.16, 2.16, 2.50, 2.83, 3.16, 3.33, 3.83, 4.50, 4.50점으로 내･외부 관측자가 영상을 평가하 였고(p<0.01), T1 영상에서 6~15일 경우 1.50, 2.16, 2.16, 2.33, 2.66, 3.33, 3.00, 3.66, 4.00, 4.16점으로 평가를 하였다(p<0.01).
결 론：교정 후 치아의 유지를 위해 남아 있는 강자성체 인공물이나 불가피하게 두･경부에 잔존하는 물질로 인해 검사에 제한 사항이 발생을 할 경우 T2 SEMAC의 경우 PES 7, T1 SEMAC의 경우는 6~8(SNR, artifact, scan time 고려시: PES 8, PES 7, PES 6)을 권고한다. 본 연구에서 제시한 최적의 T1, T2의 SEMAC PES를 참고하여 임상에 적용한다면 기존 검사법과 비교 시 영상의 질 향상에 도움이 될 것이라 판단된다.

Purpose：This study aims to determine the optimal step in reducing artifact through the measurement of the long and short axises of the axial artifact of the T1, T2 test sequence by applying the SEMAC technique on the ferromagnetic artifact remnant in the head and neck that formed after orthodontic treatment in high magnetic field (3.0T) MRI, and to apply additional techniques in actual clinical practice.
Materials and Methods：The therapeutic bolus and the dental stainless steel wire were inserted and fixed into the wrist phantom that was made by using the 3.0T MRI. The SEMAC technique was applied in the T1, T2 TSE of the pulse sequence used in this study. Measurement was repeated 10 times for each additional PES (phase encoding step), which was changed from 6 to 15. In the quantitative analysis, the major and minor axis of the left and the right part of the region where image distortion was most severe (image 18) were measured, as well as the SNR of three areas in the left and right parts, respectively, that were not affected by the artifact. Qualitative analysis was also performed by evaluating the quality of the image on a scale of 1-5 according to the image evaluation criteria set by three internal and three external evaluators.
Results：In the quantitative analysis of the artifacts of the T2 SEMAC, the change in the length ratio(%) of the artifact was measured, as the PES was adjusted as 6→7, 7→8, 8→9, and 9→10. The length on the right side was shortened by 0.11%, 0.02%, 0.10%, and 0.02% in the major axis, and by 0.19%, 0.04%, 0.22%, and 0.07% in the minor axis. On the left side, it was shortened by 0.12%, 0.02%, 0.10%, and 0.06% in the major axis, and by 0.20%, 0.09%, 0.18%, and 0.30% in the minor axis. In the quantitative analysis of the T1 images, the major axis on the right side was reduced by 0.17%, 0.01%, 0.11%, and 0.01%, and the minor axis by 0.14%, 0.01%, 0.11%, and 0.02%. The major axis of the left side was reduced by 0.20%, 0.01%, 0.09%, and 0.01%, and the minor axis by 0.13%, 0.03%, 0.11%, and 0.01% (p<0.01). The SNR measured through the elevation of the PES from 6 to 10 was 101.92, 105.25, 105.44, 104.44, and 103.47 for T2 (RT), 95.30, 98.98, 97.22, 96.61, and 95.74 for T2 (LT), 177.24, 175.50, 296.06, 299.88, and 313.71 for T1 (RT), and 159.67, 158.79, 246.75, 226.75, and 259.67 for T1 (LT). In the qualitative analysis, the internal-external evaluator assessed the T2, T1 image as 1.50, 2.16, 2.16, 2.50, 2.83, 3.16, 3.33, 3.83, 4.5, and 4.5 by using the scale of 1-5 under the SEMAC PES from 6 to 15, and 1.50, 2.16, 2.16, 2.33, 2.66, 3.33, 3.00, 3.66, 4.00, and 4.16 under the T1 image from 6 to 15, respectively.
Conclusions：We suggest testing under PES 7 with T2 SEMAC and PES 6-8 (PES 8, PES 7, and PES 6 when SNR, artifact, and scan time are considered) for T1 SEMAC in circumstances where examination is limited due to inevitable remnants in the head and neck, or ferromagnetic artifacts for maintenance of teeth following an orthodontic procedure. Applying the suggestions of this study on optimal T1 and T2 SEMAC PES in the clinical practice will improve image quality, compared to using conventional methods.

I. 서 론
 Ⅱ. 대상 및 방법
  1. 연구 대상
  2. 검사방법
  3. 영상의 분석 방법
  4. 통계적 분석 방법
 Ⅲ. 결 과
  1. 정량적 분석 결과
  2. 정성적 분석 결과
 Ⅳ. 고찰 및 결론
 References

• 이재흔(부산대학교병원 영상의학과) | Jaeheun Lee (Department of Radiology, Pusan National University Hospital)
• 장재홍(부산대학교병원 영상의학과) | Jaehong Jang (Department of Radiology, Pusan National University Hospital)
• 김태규(부산대학교병원 영상의학과) | Taegyu Kim (Department of Radiology, Pusan National University Hospital)
• 최성규(부산대학교병원 영상의학과) | Seonggyu Choi (Department of Radiology, Pusan National University Hospital)
• 김지영(부산대학교병원 영상의학과) | Jiyoung Kim (Department of Radiology, Pusan National University Hospital)
• 강남구(부산대학교병원 영상의학과) | Namgu Kang (Department of Radiology, Pusan National University Hospital) Corresponding Author