[논문]판정행렬분석을 통한 PET-MRI의 간암 진단성능 평가

김진의; 김정수; 최남길; 한재복

doi:10.5392/jkca.2017.17.11.050

판정행렬분석을 통한 PET-MRI의 간암 진단성능 평가
Evaluation of the Liver Cancer Diagnosis Function of PET-MRI Based on Decision Matrix Analysis 원문보기

한국콘텐츠학회논문지 = The Journal of the Korea Contents Association, v.17 no.11, 2017년, pp.50 - 59

김진의 (동신대학교 방사선학과) , 김정수 (동남보건대학교 방사선과) , 최남길 (동신대학교 방사선학과) , 한재복 (동신대학교 방사선학과)

초록
AI-Helper

최근 임상에서 활용하고 있는 일체형 PET-MRI의 간암 진단능력을 평가하기 위해 $2{\times}2$ 판정행렬을 이용하여 유용성을 평가하였다. 실험대상의 PET-MRI 검사 결과를 통해 간암 판정 여부 즉 비정상과 정상 판정을 받은 경우는 각각 98건, 51건 이었으며, 세포병리학적 결과가 양성과 음성 판정을 받은 경우는 각각 103건, 62건으로 나타났다. 이 중 두가지 검사에서 진양성의 경우는 95건, 위양성은 3건으로 나타났으며, 진음성은 62건, 위음성의 경우는 5건으로 분석되었다. 실험결과 PET-MRI 검사의 예민도는 95.00%, 특이도는 95.38%, 위음성률은 0.05%, 위양성률은 0.05%, 정확도는 95.15%로 분석되었다. 따라서 간암의 진단에 있어 수술 전 병기 결정이나 치료 후 재발 및 원격전이의 발견, 불분명한 원발 림프절 전이 등의 평가 등에 활용 가능성이 높을 것으로 판단되며, 특히 병리학적 검사와의 복합적 진단 및 추적검사를 통해 간암 진단을 위한 PET-MRI 임상적 유용성은 충분할 것으로 사료된다.

Abstract ▼ AI-Helper

To evaluate the capability of integrated PET-MRI, which has recently been utilized in the clinical practices, on the diagnosis of liver cancer, its utility was assessed by $2{\times}2$ decision matrix. The numbers of abnormal and normal decisions on the liver cancer were 98 and 51 cases, respectively, upon PET-MRI scan results of the subjects, and the numbers of positive and negative decisions were 103 and 62, respectively, upon cytopathologic results. Out of the two tests, 95 cases were shown as true-positive and 3 were false positive, while 62 were true negative and 5 were false negative. Upon the results of PET-MRI test, its sensitivity, specificity, false negative rate, and false positive rate were 95.00%, 95.38%, 0.05%, and 95.15%, respectively. Therefore, it is considered to have the high potential to use the determination of the stage before the surgery, detections of recurrence and remote metastasis, assessment of uncertain remote lymph node metastasis, and so on in the diagnosis of the liver cancer, and also for the clinical utility of PET-MRI to be sufficient by integrated diagnosis and follow up scan with pathological studies.

주제어

표/그림 (9)

그림 Fig. 1. PET-MRI brings the best of both worlds of PET that can diagnose liver cancer by detecting biochemical changes in the body through glucose metabolism, and MRI that can observe in high resolution soft tissue such as the liver. The integrated PET-MRI used in this study allows the artifact due to movement to be minimized as the subjects lie done to have the PET and MRI taken simultaneously. This also helps to shorten the test time which leads to better diagnosis accuracy.
표 Table 1. Conditions to acquire a PET-MRI image of the torso scan and dedication scan
그림 Fig. 2. A; SUV takes advantage of the fact that radioactive pharmacological products are absorbed by lesions such as cancer to compare the absorption rate across different body weights, heights and body surface areas of patients. B; In clinical settings, semi-quantitative approach is used where a higher level of absorption is seen as a general indicator for a more malignant tumor.
그림 Fig. 3. The data analyzed by EMR showed that the pathological examination for liver cancer through left hemi-hepatectomy on the subjects revealed masses and nodular lesions. Although there were no abnormalities discovered in the surrounding liver tissue, the size of the tumor was 3.1 × 2.8 × 2.1 cm and the degree of necrosis of the tumor was 82.7%. The pathological examination results clearly allowed for a firm diagnosis on whether the subject had liver cancer.
표 Table 2. Conditions for PET-MRI image reconstruction of the torso scan and dedication scan
그림 Fig. 4. For an evaluation on the data of the test, a relatively simple matrix can be used. The test results of the items to be compared are classified into normal or abnormal. Whether the disease is present or not is again used, along with the normal / abnormal categorization to create a matrix made of rows and columns. Applying a mathematical equation, the five indices of sensitivity, specificity, false negative ratio, false positive ratio and accuracy can be calculated. These indices reflect the data of the test.
표 Table 3. Normal and abnormal cases in the PET-MRI scan and positive and negative cases in the pathological examination using EMR.
표 Table 4. Case results of TP, FP, TN and FN analyzed by PET-MRI scan and EMR pathological tests
그림 Fig. 5. When a mathematical equation is applied to the TP, FP, TN and FN results acquired through a 2 × 2 decision matric analysis, the five indices of sensitivity, specificity, false negative ratio, false positive ratio and accuracy can be calculated. These indices do not vary according to the prevalence rate of the subjects. Therefore, in general, in a screening test such as a group health check-up, tests with high sensitivity rather than high specificity are used. Sensitivity is the value from the number of true positives divided by the number of total patients with the disease [TP/(TP+FN)], and is called as true positive ratio. Specificity is the value from the number of true negatives divided by the number of total patients without the disease [TN/(FP+TN)], and is called as true negative ratio. False negative ratio is the value from the number of false negatives divided by the number of total patients with the disease [FN/(TP+FN)], and the sum of true positive ratio and false negative ratio is always '1.' False positive ratio is the value from the number of false positives divided by the number of total patients without the disease [FP/(TN+FP)], and the sum of true negative ratio and false positive ratio is always '1.' Accuracy is the value from the number of true positives and true negatives divided by the number of total patients in decision matrix.

AI 본문요약
AI-Helper

* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.

제안 방법

After reconstruction the scanned results through iterative method, an interest area for the lesion was set and the SUV (standard uptake value) measured for quantitative analysis. The data of PET-MRI was then analyzed.
The PET-MRI scan protocol for liver cancer as dictated by the nuclear medicine department of SNUH was applied, and thus the subjects underwent a torso scan to check the whole body. After the test an in-depth analysis of the liver was done with a liver dedication scan that followed immediately afterwards. At this point in time, Body TIM (total imaging matrix) coil was attached and VIBE (volumetric interpolated breath-hold examination)-Dixon (TR, repetition time; 3.
SNUH (seoul national university hospital) frequently uses PET-MRI to diagnose liver cancer but there are no specific data on the practical benefits of the diagnosis method. As such, this study compares the PET-MRI imaging results of liver cancer patients with pathological opinions and pathological examination results to evaluate the practical benefits of PET-MRI scans for liver cancer patients in a clinical setting.
For an analysis of the pathological result, the hospital’s comprehensive medical information system (EMR; electronic medical record, bestcare) was used to review the pathological examination opinion[Fig. 3].
The PET-MRI scan protocol for liver cancer as dictated by the nuclear medicine department of SNUH was applied, and thus the subjects underwent a torso scan to check the whole body. After the test an in-depth analysis of the liver was done with a liver dedication scan that followed immediately afterwards.
The diagnosis information of the patient thus acquired was entered into a 2 × 2 decision matrix to categorize them as TP (true positive), TN (true negative), FP (false positive) and FN (false negative).
PET-MRI brings the best of both worlds of PET that can diagnose liver cancer by detecting biochemical changes in the body through glucose metabolism, and MRI that can observe in high resolution soft tissue such as the liver. The integrated PET-MRI used in this study allows the artifact due to movement to be minimized as the subjects lie done to have the PET and MRI taken simultaneously. This also helps to shorten the test time which leads to better diagnosis accuracy.
This study analyzed a relatively simple 2 × 2 decision matrix for the experiment data for analysis.
This is why a prospective study using a larger patient group is needed. This study was conducted on patients and should be compared against objective data using a phantom, but such comparison was not applied, making the study limited. Therefore, many clinical experiments are needed in order to generalize the study findings.
False negative ratio is the percentage of cases where the diagnosis is given as negative even though there is liver cancer, while false positive ratio is the percentage of cases where the diagnosis is given as positive even though there is no liver cancer. Using the five indices concluded through the data analysis, the diagnosis function of the PET-MRI test can be evaluated. The experiment results show sensitivity to be 95.

대상 데이터

The integrated PET-MRI equipment BiographTM mMR 3T (SIEMENS, Munich) was used for the experiment[Fig. 1].
This study is limited in that it has a small number of subjects, 165 subjects in total. This leads to the small number of patients with a specific illness to be over the mean, leading to a relatively smaller size of the patient group.

참고문헌 (32)

Ahmedin Jemal, Freddie Bray, Melissa M. Jacques Ferlay, Elizabeth Ward, and David Forman, "Global cancer statistics," Ca cancer J Clin., Vol.61, pp.69-90, 2011.

상세보기
K. W. Jung, Y. J. Won, C. M. Oh, H. J. Kong, D. H. Lee, and K. H. Lee, "Cancer statistics in korea: incidence, mortality, survival, and prevalence in 2014," Cancer research and Treatment, Vol.49, No.2, pp.292-305, 2017.

상세보기
Ran Xu Zhu, Wai-Kay Seto, Ching-Lung Lai, Man-Fung Yuen, M. J. Daly, and R. E. Henly, "Epidemiology of hepatocellular carcinoma in the asia-pacific region," Gut Liver, Vol.10, No.3, pp.332-339, 2016.
S. K. Yoon and H. G. Chun, "Status of hepatocellular carcinoma in south korea," Chin Clin Oncol, Vol.2, No.4, p.39, 2013.

상세보기
E. Y. Lee, T. T. Xuan Mai, Y. J. Chang, and M. R. Ki, "Trends of liver cancer and its major risk factors in korea," Epidemiol Health, Vol.37, p.e2015016, 2015.

상세보기
Robert G. Gish, Richard S. Finn, and Jorge A. Marrero, "Extending survival with the use of targeted therapy in the treatment of hepatocellular carcinoma," Gastroenterol Hepatol, Vol.9, No.4, suppl.2, pp.1-24, 2013.
Josep M. Llovet, Maria Isabel Real, Xavier Montana, Ramon Planas, Susana Coll, John Aponte, Carmen Ayuso, Margarita Sala, Jordi Muchart, Ricard Sola, Joan Rodes, and Jordi Bruix, "Arterial embolisation or chemoembolisation versus symptomatic," Lancet, Vol.359, pp.1734-1739, 2002.

상세보기
Gerald D. Dodd, Michael C. Soulen, Robert A. Kane, Tito Livraghi, William R. Lees, Yasuyuki Yamashita, Alison R. Gillams, Okkes I. Karahan, and H. C. Rhim, "Minimally invasive treatment of malignant hepatic tumors: at the threshold of a major breakthrough," Radiographics, Vol.20, pp.9-27, 2000.

상세보기
D. G. Farmer, M. H. Rosove, A. Shaked, and R W. Busuttil, "Current treatment modalities for hepatocellular carcinoma," Ann Surg, Vol.219, pp.236-247, 1994.

상세보기
F. F. Haddad, W. C. Chapman, J. K. Wright, T. K. Blair, and C. W. Pinson, "Clinical experience with cryosurgery for advanced hepatobiliary tumors," J Surg Res., Vol.75, pp.103-108, 1998.

상세보기
Eric C. Ehman, Geoffrey B. Johnson, Javier E. Villanueva-Meyer, S. M. Cha, Andrew Palmera Leynes, Peder Eric Zufall Larson, and Thomas A. Hope, "PET/MRI: Where might it replace PET/CT?," J Magn Reson Imaging, doi:10.1002/jmri.25711, 2017.

상세보기
Felix Nensa, Karsten Beiderwellen, Philipp Heusch, and Axel Wetter, "Clinical applications of PET/MRI: current status and future perspectives," Diagn Interv Radiol, Vol.20, No.5, pp.438-447, 2014.

상세보기
W. Atkinson, C. Catana, J. S. Abramson, G. Arabasz, S. McDermott, O. Catalano, V. Muse, M. A. Blake, J. Barnes, M. Shelly, E. Hochberg, B. R. Rosen, and A. R. Guimaraes, "Hybrid FDG-PET/MR compared to FDG-PET/CT in adult lymphoma patients," Abdom Radiol, Vol.41, pp.1338-4138, 2016.

상세보기
C. M. Tempany, J. Jayender, T. Kapur, T. Kapur, R. Bueno, A. Golby, N. Agar, and F. A. Jolesz, "Multimodal imaging for improved diagnosis and treatment of cancers," Cancer, Vol.121, pp.817-827, 2015.

상세보기
Toshio Wagai, "Studies on the foundation and development of diagnostic ultrasound," Proc Jpn Acad Ser B Phys Biol Sci., Vol.83, No.8, pp.256-265, 2007.

상세보기
A. Samarin, M. Hullner, M. A. Queiroz, P. Stolzmann, I. A. Burger, G. Von Schulthess, and P. Veit Haibach, " $^{18}F$ -FDG-PET/MR increases diagnostic confidence in detection of bone metastases compared with $^{18}F$ -FDG-PET/CT," Nucl Med Commun., Vol.36, No.12, pp.1165-1173, 2015.

상세보기
K. Beiderwellen, M. Huebner, P. Heusch, J. Grueneisen, V. Ruhlmann, F. Nensa, H. Kuehl, L. Umutlu, S. Rosenbaum-Krumme, and T. C. Lauenstein, "Whole-body [ $^{18}F$ ]FDG PET/MRI vs. PET/CT in the assessment of bone lesions in oncological patients: initial results," Eur Radiol., Vol.24, No.8, pp.2023-2030, 2014.

상세보기
K. Beiderwellen, L. Geraldo, V. Ruhlmann, P. Heusch, B. Gomez, F. Nensa, L. Umutlu, and T. C. Lauenstein, "Accuracy of [ $^{18}F$ ]FDG PET/MRI for the detection of liver metastases," PLoS ONE, Vol.10, No.9, p.e0137285, 2015.

상세보기
Philipp Heusch, Felix Nensa, Benedikt Schaarschmidt, Rupika Sivanesapillai, Karsten Beiderwellen, Benedikt Gomez, Jens Kohler, Henning Reis, Verena Ruhlmann, and Christian Buchbender, "Diagnostic accuracy of whole-body PET/MRI and whole-body PET/CT for TNM staging in oncolog," Eur J Nucl Med Mol Imaging, Vol.42, No.1, pp.42-48, 2015.

상세보기
Ali Civelek, Anshuma Rana, Ashkan Malayeri, Justin Rodante, Amit Dey, Nehal Mehta, and David Bluemke, "PET-MRI derived 18F-FDG liver SUV metrics: effects of hepatic steatosis and comparison with PET-CT," J Nucl Med., Vol.58, No.1, p.455, 2017.
Kathy Schilling, Deepa Narayanan, Judith E. Kalinyak, Juliette The, Maria Victoria Velasquez, Simone Kahn, Matthew Saady, Ravinder Mahal, and Larraine Chrystal, "Positron emission mammography in breast cancer presurgical planning: comparisons with magnetic resonance imaging," Eur J Nucl Med Mol Imaging, Vol.38, No.1, pp.23-36, 2011.

상세보기
Ahmet Mert, Niyazi Kilic, Erdem Bilgili, and Aydin Akan, "Breast cancer detection with reduced feature set," Comput Math Methods Med., doi:10.1155/2015/265138, 2015.

상세보기
A. Marcano-Cedeno, Paloma Chausa, Alejandro Garcia, Cesar Caceres, Josep M. Tormos, and Enrique J. Gomez, "Data mining applied to the cognitive rehabilitation of patients with acquired brain injury," Exp Sys Appl., Vol.40, No.4, pp.1054-1060, 2013.

상세보기
Dushyant V. Sahani and Sanjeeva P. Kalva, "Imaging the liver," The Oncologist, Vol.9, No.4, pp.385-397, 2004.

상세보기
Claudio Spick, Ken Herrmann, and Johannes Czernin, " $^{18}F$ -FDG PET/CT and PET/MRI perform equally well in cancer: evidence from studies on more than 2,300 Patients," J Nucl Med, Vol.57, No.3, pp.420-430, 2016.

상세보기
Ruben Hernaez, Mariana Lazo, Susanne Bonekamp, Ihab Kamel, Frederick L. Brancati, Eliseo Guallar, and Jeanne M. Clark, "Diagnostic accuracy and reliability of ultrasonography for the detection of fatty liver: a meta-analysis," Hepatology, Vol.54, No.3, pp.1082-1090, 2011.
Eldad S. bialecki and Adrian M. Di Bisceglie, "Diagnosis of hepatocellular carcinoma," HPB(Oxford), Vol.7, No.1, pp.26-34, 2005.

상세보기
A. Abbasi, A. R. Bhutto, N. Butt, and S. M. Munir, "Corelation of serum alpha fetoprotein and tumor size in hepatocellular carcinoma," J Pak Med Assoc, Vol.62, No.1, pp.33-36, 2012.
Marialuigina Fruscella, Antonio Ponzetto, Annalisa Crema, and Guido Carloni, "The extraordinary progress in very early cancer diagnosis and personalized therapy: the role of oncomarkers and nanotechnology," J Nanotech, doi:10.1155/2016/3020361, 2016.

상세보기
Bing Bai, Quanzheng Li, and Richard M. Leahy, "MR guided PET image reconstruction," Semin Nucl Med., Vol.43, No.1, pp.30-44, 2014.
Andrew J. Vickers, "Decision analysis for the evaluation of diagnostic tests, prediction models and molecular markers," Am Stat., Vol.62, No.4, pp.314-320, 2008.

상세보기
Peter R. Killeen, "Beyond statistical inference: A decision theory for science," Psychon Bull Rev., Vol.13, No.4, pp.549-562, 2006.

상세보기

저자의 다른 논문 :

표제어: PCR

동의어: Packet Collision Rate

용어 설명 출처 목록 (6)

용어 설명: PCR은 세균 특이성이 있는 primer를 이용하여 적은 수의 세균이 있을지라도 쉽게 검출할 수 있는 유용한 방법이며, 이를 이용하여 구강 내 치면세균막이나 타액에서 직접 세균을 검출할 수 있게 되었다[8].

내보내기 구분	파일저장 인쇄 메일전송
구성항목	기본정보 상세정보 관리번호, 논문명, 저널/프로시딩명, 저자 , 발행년, 권, 호, 시작페이지, 끝페이지, 발행기관 관리번호, 논문명, 대등논문명, 저자 , 저널/프로시딩명, 발행기관, 발행년, 발행언어, 권, 호, 시작페이지, 끝페이지, ISBN, ISSN, 주제분야, 키워드, 초록(한글), 초록(영문), 저자(소속기관)
저장형식	Text(ASCII format) Excel format RefWorks Direct Export RIS format (for Reference Manager, ProCite, EndNote), Scholar's Aids, Mendeley
메일정보	받는사람 (필수) @ 보내는사람 (선택) @ 제목 내용 KISTI 검색결과 이메일 서비스
안내	총 건의 자료가 검색되었습니다. 다운받으실 자료의 인덱스를 입력하세요. (1-10,000) 검색결과의 순서대로 최대 10,000건 까지 다운로드가 가능합니다. 데이타가 많을 경우 속도가 느려질 수 있습니다.(최대 2~3분 소요) 다운로드 파일은 UTF-8 형태로 저장됩니다. 파일의 내용이 제대로 보이지 않을실 때는 웹브라우저 상단의 보기 -> 인코딩 -> 자동선택 여부를 확인하십시오. ~ Text(ASCII format) Excel format

연합인증