Quantitative imaging of hepatic cirrhosis on abdominal CT images Poster No.: C-0556 Congress: ECR 2013 Type: Authors: Keywords: DOI: Scientific Exhibit S. Kido, A. Nakamura, Y. Hirano; Ube/JP Cirrhosis, Computer Applications-Detection, diagnosis, CAD, CT, Liver, Computer applications, Abdomen 10.1594/ecr2013/C-0556 Any information contained in this pdf file is automatically generated from digital material submitted to EPOS by third parties in the form of scientific presentations. References to any names, marks, products, or services of third parties or hypertext links to thirdparty sites or information are provided solely as a convenience to you and do not in any way constitute or imply ECR's endorsement, sponsorship or recommendation of the third party, information, product or service. ECR is not responsible for the content of these pages and does not make any representations regarding the content or accuracy of material in this file. As per copyright regulations, any unauthorised use of the material or parts thereof as well as commercial reproduction or multiple distribution by any traditional or electronically based reproduction/publication method ist strictly prohibited. You agree to defend, indemnify, and hold ECR harmless from and against any and all claims, damages, costs, and expenses, including attorneys' fees, arising from or related to your use of these pages. Please note: Links to movies, ppt slideshows and any other multimedia files are not available in the pdf version of presentations. www.myesr.org Page 1 of 9
Purpose Assessment of hepatic cirrhosis is a very important diagnostic and prognostic evaluation in chronic hepatic diseases. This assessment of hepatic cirrhosis involves a semiquantitative pathologic scoring system of a liver biopsy specimen. However, liver biopsies are invasive and carry risks of complications such as hemorrhage and infection. Currently, abdominal CT scan is widely used for diagnostic and prognostic evaluation for chronic hepatic diseases. Abdominal CT images can provide quantitative information about hepatic cirrhosis. However, assessments of hepatic cirrhosis among radiologists are wildly different, because they do not fully use such quantitative information. So, the purpose of our study was to quantify hepatic cirrhosis on abdominal CT images using Computer-aided diagnosis (CAD) algorithm by use of a texture analysis method and evaluate the usefulness of our algorithm compared with pathological diagnosis. Methods and Materials In our CAD scheme, hepatic regions were extracted from abdominal CT volume data and texture features based on CT values were calculated from extracted hepatic regions on abdominal CT volume data. Each case was classified based on these features by use of a Support Vector Machine. A. CAD algorithms In the first step, we determined a seed area for hepatic extraction. In the next step, we extracted a fine hepatic region based on this seed area by use of a level-set method [1] (Fig.2). For the extracted hepatic regions, we calculated the texture features based on CT values. We used a grey-level co-occurrence matrix and a grey-level-run length matrix for feature calculation of hepatic cirrhosis. Each case was classified based on these features by use of a Support Vector Machine. B. Patients We used 20 patients (16 Males, 4 Females, Mean age was 61.7 ± 11.8 years) with chronic hepatic diseases who underwent non-enhanced CT scans using a multi-detector row CT scanner. In this analysis, we did not use enhanced CT images, because nonenhanced CT scans are performed in routine CT studies in Japan. All cases were assessed pathologically based on a new Inuyama classification of cirrhotic hepatitis [2] (Fig.3). In this classification, chronic hepatic diseases are classified according to the degree of fibrosis (F) as follows; F0 (no fibrosis), F1 (portal fibrosis widening), F2 (portal fibrosis widening with bridging fibrosis), F3 (bridging fibrosis plus lobular distortion), Page 2 of 9
and F4 (liver cirrhosis). And, also chronic hepatic diseases are classified based on the degree of inflammation of lymphocytes and necrosis of hepatocytes. The activities (A) are classified from A0 to A3 as follows; A0 (no necro-inflammatory reaction), A1 (mild necro-inflammatory reaction), A2 (moderate necro-inflammatory reaction), and A3 (severe necro-inflammatory reaction). Our cases included no F0 case, 10 mild cases (F1: n=2, F2: n=8), and 10 sever cases (F3: n=3, F4: n=7), and also our cases included no A0 cases, 8 A1 cases, 12 A2 cases, and no A3 cases. Images for this section: Fig. 1: In our CAD scheme, hepatic regions were extracted from abdominal CT volume data and texture features based on CT values were calculated from extracted hepatic regions on abdominal CT volume data, and each case was classified based on these features by use of a Support Vector Machine. Page 3 of 9
Fig. 2: In the first step, we determined a seed area for hepatic extraction. In the next step, we extracted a fine hepatic region based on this seed area by use of a level-set method. Page 4 of 9
Fig. 3 Page 5 of 9
Results The correct classification rate of fibrosis for all cases was 15/20 (75.0%). Number of correct classification cases for each category was 0/2 (F1), 7/8 (F2), 2/3 (F3), and 6/7 (F4) (Table1). The correct classification rate of necro-inflammatory changes for all cases was 18/20 (90%). Number of correct classification cases for each category was 7/8 (A1) and 11/13 (A2) (Table2) Images for this section: Table 1 Page 6 of 9
Table 2 Page 7 of 9
Conclusion The texture features obtained from hepatic regions with chronic hepatic diseases were found to reflect the degree of pathological hepatic cirrhosis. Our CAD algorithm has potential usefulness for diagnosis of hepatic cirrhosis non-invasively. References [1] Kido S, Ohga Y. Subtraction Method for Detection of Hepatic Tumors with Segmentation using Level Set and Registration using Phase Only Correlation in Multiphase Multi-detector Row CT Images. Medical Imaging and Information Sciences 1991; 25: 86-89. [2] Ichida F, Tsuji T, Omata M, et al. New Inuyama classification; new criteria for histological assessment of chronic hepatitis. International hepatology communications 1996; 6: 112-119. [3] Committee on Chronic Hepatitis, Japanese Society of Hepatology. Chronic hepatitis, Proc 1 st Inuyama symp Publicaion Committee of Inuyama Symposium, 1967: 1-50, 165-170. (In Japanese). [4] De Groote J, Desmet VJ, Gedigk P, et al. A classification of chronic hepatitis. Lancet 1968: 2: 626-628. [5] Committee of Inuyama Symposium. A new classification of chronic hepatitis. Proc 11 th Inuyama Symp Chugai Igakusha, Tokyo, 1983: 1-83. (In Japanese) [6] Bianchi L, De Groote J, Desmet VJ, et al. (International Group). Acute and chronic hepatitis revisited. Lancet 1977; 2: 914-919. Personal Information Shoji Kido, MD, PhD Computer-aided Diagnosis and Biomedical Imaging Research Biomedical Engineering, Applied Medical Engineering Science Graduate School of Medicine, Yamaguchi University Page 8 of 9
E-mail: kido@ai.csse.yamaguchi-u.ac.jp Page 9 of 9