Characterization of breast lesions using diffusion kurtosis model-based imaging: An initial experience

Article type: Research Article

Authors: Zhou, Wei-Ping^{a; 1} | Zan, Xing-You^{b; 1} | Hu, Xiao-Yun^{a; 1} | Liu, Xiao^c | Sudarshan, Sanjeev Kumar Pitla^d | Yang, Shu-Dong^e | Guo, Yu-Jiang^{c; *} | Fang, Xiang-Ming^{a; *}

Affiliations: [a] Department of Radiology, Wuxi People’s Hospital, Nanjing Medical University, Jiangsu, China | [b] Department of Ultrasound, Wuxi People’s Hospital, Nanjing Medical University, Jiangsu, China | [c] Department of Thyroid Breast Surgery, Wuxi People’s Hospital, Nanjing Medical University, Jiangsu, China | [d] Department of Medical Imaging, Parexel International Corporation, Billerica, Massachusetts, USA | [e] Department of Pathology, Wuxi People’s Hospital, Nanjing Medical University, Jiangsu, China

Correspondence: [*] Corresponding authors: Xiang-Ming Fang, Department of Radiology, Wuxi People’s Hospital, Nanjing Medical University, Jiangsu, China. Tel.: +86 13861779030; E-mail: xiangming fang@njmu.edu.cn and Yu-Jiang Guo, Department of Thyroid Breast Surgery, Wuxi People’s Hospital, Nanjing Medical University, Jiangsu, China. Tel.: +86 13358112052; E-mail: wxwkgyj2007@126.com.

Note: [1] Co-author and contributed equally authors for the first author.

Abstract: OBJECTIVE:To investigate the characterization of breast lesions using diffusion kurtosis model-based imaging. METHODS:This prospective study included 120 consecutive patients underwent preoperative DCE-MRI examinations and multi-b-value diffusion-weighted imaging (DWI). Among them, 88 malignant lesions and 44 benign lesions were detected, 56 normal fibroglandular breast tissue were selected as normal control. Conventional apparent diffusion coefficient (ADC), DKI-based parameters mean kurtosis (MK) and mean diffusivity (MD) were analyzed by lesions types and histological subtypes using one-way ANOVA and receiver operating characteristic (ROC) curve. RESULTS:(1) The malignant group showed significantly lower ADC and MD (1.07±0.32×10-3 mm2/s and 1.30±0.40×10-3 mm2/s, respectively) and higher MK (0.87±0.18) than those in the benign group (1.29±0.26×10-3 mm2/s, 1.62±0.31×10-3 mm2/s and 0.67±0.18) and control group (1.67±0.33×10-3 mm2/s, 2.24±0.28×10-3 mm2/s and 0.52±0.08) with all P < 0.001. (2) Areas under ROC curve (AUC) for diagnosing malignant lesions were 0.936 for MD, 0.911 for MK and 0.897 for ADC, respectively. AUC for MD was significantly higher than that for ADC (P = 0.015). The optimal cut-off value, sensitivity, specificity, positive predictive value, negative predictive value and accuracy were as follow: ADC = 1.18×10-3mm2/s, 78.3%, 93.2%, 81.2%, 81.6%, 81.4%; MD = 1.48×10-3mm2/s, 82.2%, 98.3%, 84.4%, 87.8%, 86.2%; MK = 0.78, 91.5%, 85.3%, 89.0%, 85.8%, 87.2%. (3) Invasive ductal carcinoma (IDC), ductal carcinoma in situ (DCIS) and mucinous adenocarcinoma also showed significant differences among ADC, MD and MK (with P < 0.05). CONCLUSIONS:MR-DKI parameters enable to improve breast lesion characterization and have diagnostic potential applying to different pathological subtypes of breast cancers.

Keywords: Diffusion weighted imaging, diffusion kurtosis imaging, breast lesions, pathological subtypes

DOI: 10.3233/XST-190590

Journal: Journal of X-Ray Science and Technology, vol. 28, no. 1, pp. 157-169, 2020