690 AJR:202, March 2014 tive rate, and concerns have been expressed regarding its operator dependence [2–4]. Molecular imaging technologies have been developed recently to circumvent these limitations. Breast-specific gamma imaging (BSGI), also referred to as “molecular breast imaging,” has been improved significantly in recent years with the development of breast- optimized, high-resolution, small-FOV gam- ma camera designs [5]. Unlike mammog- raphy and ultrasound, BSGI is a functional imaging examination that reflects the bio- chemical and physiologic characteristics of tumors. In particular, cellular mitochondri- al density can be measured using 99m Tc-me- thoxyisobutylisonitrile (MIBI) as a tracer; high cytoplasmic mitochondrial density is typical of hyperproliferative cell types and not of benign pathologic entities [6]. Complementary Role of Semiquantitative Analysis of Breast-Specifc Gamma Imaging in the Diagnosis of Breast Cancer Kyung Sik Park 1 Hyun Woo Chung 2 Young Bum Yoo 1 Jung-Hyun Yang 1 Nami Choi 3 Young So 2 Park KS, Chung HW, Yoo YB, Yang JH, Choi N, So Y 1 Department of Surgery, Konkuk University Medical Center, Konkuk University School of Medicine, Seoul, Korea. 2 Department of Nuclear Medicine, Konkuk University Medical Center, Research Institute of Biomedical Science, Konkuk University School of Medicine, 120-1 Neungdong-ro (Hwayang-dong), Gwangjin-gu, Seoul 143-729, Korea. Address correspondence to H. W. Chung (hwchung@kuh.ac.kr). 3 Department of Radiology, Konkuk University Medical Center, Konkuk University School of Medicine, Seoul, Korea. Women’s Imaging • Original Research AJR 2014; 202:690–695 0361–803X/14/2023–690 © American Roentgen Ray Society T he global incidence and mortality of breast cancer have increased steadily in the past decades, and breast cancer is recognized as an important health problem for women [1]. Mammography and ultrasound are commonly used anatomic imaging procedures to detect breast cancer, but they have several limitations. The sensitivity of mammography for breast cancer detection decreases substantially if the breast parenchyma is dense. In addition, false- positive diagnoses based on mammography re- sult in many benign findings at biopsy. Ultra- sound, the most commonly used adjunct breast imaging technique, can depict small node-neg- ative breast cancers, which increases the prob- ability of cancer detection, especially in wom- en with mammographically dense breast tissue; however, ultrasound also has a high false-posi- Keywords: breast cancer, breast-specifc gamma imaging (BSGI), mammography, semiquantitative analysis, ultrasound DOI:10.2214/AJR.13.11324 Received May 27, 2013; accepted after revision July 21, 2013. Presented in part at the 2012 annual meeting of the Society of Nuclear Medicine and Molecular Imaging, Miami, FL. This work was supported by Konkuk University. OBJECTIVE. We investigated whether the interpretation of breast-specific gamma imaging (BSGI) with visual and semiquantitative analyses can improve the diagnosis of breast cancer. MATERIALS AND METHODS. The records of 114 women (mean age ± SD, 49.6 ± 9.8 years) who underwent BSGI, mammography, and ultrasound to evaluate a breast lesion or lesions were reviewed retrospectively. The breast lesions identified with BSGI were com- pared with those identified with mammography and ultrasound. BSGI was first interpreted visually, and then a semiquantitative analysis was performed. For the semiquantitative analy- sis, the uptake ratio for each breast lesion was calculated by dividing the tumor uptake by the contralateral normal breast uptake. RESULTS. Four of the 114 patients had two breast lesions, so a total of 118 breast lesions (42 malignant lesions and 76 benign lesions) were evaluated. A BSGI uptake ratio cutoff of 1.5, with values less than 1.5 indicating negative for cancer, as determined by receiver oper- ating characteristic curve analysis of our data (area under curve, 0.874), was used for semi- quantitative analysis. The sensitivity and specificity of BSGI with visual analysis alone for assessing malignant breast lesions were 76.2% (32/42) and 81.6% (62/76), respectively. For BSGI with visual and semiquantitative analyses, the sensitivity and specificity were 76.2% (32/42) and 92.1% (70/76), respectively. The sensitivity and specificity for mammography were 57.1% (24/42) and 81.6% (62/76), respectively. For ultrasound, the respective values were 97.6% (41/42) and 61.8% (47/76). BSGI with visual and semiquantitative analyses had a significantly higher specificity than BSGI with visual analysis alone, mammography, and ultrasound (all, p < 0.01). CONCLUSION. Semiquantitative analysis of BSGI with visual interpretation may be a useful complementary method for evaluating malignant breast lesions. Park et al. The Role of BSGI in Breast Cancer Diagnosis Women’s Imaging Original Research Downloaded from www.ajronline.org by PJERIN LULI on 07/07/14 from IP address 162.83.115.135. Copyright ARRS. For personal use only; all rights reserved