Ming-Che Lee1,2,3, Keh-Shih Chuang1, Mu-Kuan Chen4, Chi-Kuang Liu2, Kwo-Whei Lee2, Hui-Yu Tsai5,6,7, Hsin-Hon Lin1,6. 1. 1 Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu, Taiwan. 2. 2 Department of Medical Imaging, Changhua Christian Hospital, Changhua, Taiwan. 3. 3 Department of Medical Imaging and Radiological Science, Central Taiwan University of Science and Technology, Taichung, Taiwan. 4. 4 Superintendent's Office, Changhua Christian Hospital, Changhua, Taiwan. 5. 5 Department of Medical Imaging and Radiological Sciences, College of Medicine, Chang Gung University, Taoyuan, Taiwan. 6. 6 Medical Physics Research Center, Institute for Radiological Research, Chang Gung University/Chang Gung Memorial Hospital, Taoyuan, Taiwan. 7. 7 Department of Medical Imaging and Intervention, Chang Gung Memorial Hospital Linkou, College of Medicine, Chang Gung University, Taoyuan, Taiwan.
Abstract
OBJECTIVE: The present study evaluated and analyzed apparent diffusion coefficients (ADCs) from partitions through a fuzzy C-means (FCM) technique for distinguishing nodal metastasis in head and neck cancer. METHODS: MRI studies of 169 lymph node lesions, dissected from 22 patients with a histopathologically confirmed lymph node status, were analyzed using in-house software developed using MATLAB(®) (The MathWorks(®) Inc., Natick, MA). A radiologist manually contoured the lesions, and ADCs for each lesion were divided into two (low and high) and three (low, intermediate and high) partitions by using the FCM clustering algorithm. RESULTS: The results showed that the low-value ADC clusters were more sensitive (95.7%) in distinguishing malignant from benign lesions than the whole-lesion mean ADC values (78.3%), while retaining a high specificity (approximately 90%). Moreover, receiver-operating characteristic curves demonstrated that the low-value ADC clusters used as a predictor of malignancy for lymph nodes could achieve a higher area under the curve (0.949 and 0.944 for two and three partitions, respectively). CONCLUSION: The segmentation by ADC values of lesions through the FCM technique enables the efficient characterization of the lymph node pathology and can help distinguish malignant from benign lymph nodes. ADVANCES IN KNOWLEDGE: Tumour heterogeneity may degrade the prediction of metastatic lymph nodes that involves using mean region-of-interest ADC values. The clustering of ADC values in lesions by using FCM can improve the diagnostic accuracy of nodal metastasis and reduce interreader variance.
OBJECTIVE: The present study evaluated and analyzed apparent diffusion coefficients (ADCs) from partitions through a fuzzy C-means (FCM) technique for distinguishing nodal metastasis in head and neck cancer. METHODS: MRI studies of 169 lymph node lesions, dissected from 22 patients with a histopathologically confirmed lymph node status, were analyzed using in-house software developed using MATLAB(®) (The MathWorks(®) Inc., Natick, MA). A radiologist manually contoured the lesions, and ADCs for each lesion were divided into two (low and high) and three (low, intermediate and high) partitions by using the FCM clustering algorithm. RESULTS: The results showed that the low-value ADC clusters were more sensitive (95.7%) in distinguishing malignant from benign lesions than the whole-lesion mean ADC values (78.3%), while retaining a high specificity (approximately 90%). Moreover, receiver-operating characteristic curves demonstrated that the low-value ADC clusters used as a predictor of malignancy for lymph nodes could achieve a higher area under the curve (0.949 and 0.944 for two and three partitions, respectively). CONCLUSION: The segmentation by ADC values of lesions through the FCM technique enables the efficient characterization of the lymph node pathology and can help distinguish malignant from benign lymph nodes. ADVANCES IN KNOWLEDGE: Tumour heterogeneity may degrade the prediction of metastatic lymph nodes that involves using mean region-of-interest ADC values. The clustering of ADC values in lesions by using FCM can improve the diagnostic accuracy of nodal metastasis and reduce interreader variance.
Authors: Anwar R Padhani; Andreas Makris; Peter Gall; David J Collins; Nina Tunariu; Johann S de Bono Journal: J Magn Reson Imaging Date: 2014-02-10 Impact factor: 4.813
Authors: J Wang; S Takashima; F Takayama; S Kawakami; A Saito; T Matsushita; M Momose; T Ishiyama Journal: Radiology Date: 2001-09 Impact factor: 11.105