Modified Dixon fat-water imaging for detecting intravertebral fracture line in the benign and malignant vertebral compression fractures : usefulness of fat, water, in phase and opposed phase imaging

Abstract

이 연구는 척추체 내부의 골절선으로 양성 척추 압박 골절과 악성 척추 압박 골절을 감별하는 것에 modified Dixon 자기공명영상이 진단적 역할을 하는가를 평가하는 것이다. 60명의 환자가 초기 명부에 올랐으며, 이들은 2014년 4월부터 2016년 8월까지 시상면 T2강조 two-point modified Dixon 터보스핀에코영상에서 척추압박골절을 진단받은 자였다. 제외 기준을 적용한 후 37명의 환자가 남았으며, 이들은 임상적 병력, 조직검사, 자기공명영상 검사 전후 5개월 이내의 양전자 방출 컴퓨터 단층 촬영 결과에 따라 양성 척추 압박 골절군과 악성 척추 압박 골절군으로 분류되었다. 두 명의 근골격 영상의학 전문의가 시상면 T2 강조 two-point modified Dixon 터보스핀에코영상을 판독하였으며, 척추체 내부의 골절선 유무를 독립적으로 평가하였다. 일반화된 산정 균등화 및 가중치 최소 자승법을 이용하여 water only, fat only, in phase, opposed phase 각각에서의 골절선 발견의 민감도, 특이도, 양성예측도, 음성예측도, 정확도를 산출하였으며, 관찰자 간 신뢰도는 카파 계수를 이용하여 평가하였다. 총 73개/37명의 척추 압박 골절 중 32개/14명이 양성 압박 골절이었고, 나머지 41개/23명은 악성 압박 골절이었다. Opposed phase 의 민감도는 water only 영상, fat only 영상, in phase 영상보다 통계적으로 유의하게 높았으며, fat only 영상의 특이도는 water only 영상, in phase 영상, opposed phase 영상과 비교시 통계적으로 유의하게 높았다. 관찰자 간 신뢰도는 substantial-good 이었다. 결론적으로, two-point modified Dixon 자기공명영상을 이용할 시, 기존 영상기법보다 척추체 내부의 골절선을 더 잘 발견하게 되고, 이를 통해 양성 압박 골절과 악성 압박 골절의 감별에 도움을 받을 수 있다는 것이 증명되었다.

The purpose of this study is to evaluate the diagnostic role of the modified Dixon magnetic resonance imaging for differentiation between benign vertebral compression fractures and malignant vertebral compression fractures focusing on the intravertebral fracture line. Patients were initially enrolled from April 2014 to August 2016. Sixty patients with a radiological report of vertebral compression fractures were collected. All sixty patients underwent sagittal T2-weighted two-point modified Dixon turbo spin echo sequences. After application of the exclusion criteria, 37 patients were finally enrolled. The clinical history, pathologic confirmation, and positron emission tomography-computed tomography within 5 months from the time of modified Dixon magnetic resonance imaging examination were used as a reference standard to differentiate between benign vertebral compression fractures and malignant vertebral compression fractures. Two radiologists independently evaluated the presence or absence of the fracture lines on sagittal T2-weighted two-point modified Dixon turbo spin echo sequences. Sensitivities, specificities, positive predictive values, negative predictive values, and accuracies in detecting fracture line on water only image, fat only image, in phase image, and opposed phase image were determined with the generalized estimating equation method or weighted least square. The interobserver reliability was evaluated by kappa value. A total of 73 vertebral body collapse were found in 37 patients : study group was composed of 14 patients with 32 benign vertebral compression fractures and 23 patients with 41 malignant vertebral fractures. The sensitivity of opposed phase image was higher than that of water only image, fat only image, and in phase image in the post hoc Bonferroni test with p < 0.05. The specificity of fat only image was higher than that of water only image, in phase image, and opposed phase image (p <0.0001). The positive predictive value of fat only image was significantly higher than that of water only image, in phase image, and opposed phase image (p value less than 0.05). Interobserver reliability was substantial to good. In conclusion, modified Dixon magnetic resonance imaging may be a helpful option in patients to discriminate benign vertebral compression fractures andma lignant vertebral compression fractures.