Powered by
06:14 CET
EM 3 - A glance of China through images
Imaging Methods Artificial Intelligence & Machine Learning
Sunday, March 4, 10:30 - 12:00
Room: B
Moderators: B. Hamm (Berlin/DE), Z. Y. Jin (Beijing/CN)
Z. Y. Jin; Beijing/CN
Learning Objectives

1. To give an overview of the current situation and trend of radiology in the mainland of China.
2. To show a few examples of the diversity of clinical radiological researches in China.
3. To introduce the special focuses of Chinese radiological society.


The Chinese radiological system is continuously innovating. Fifteen years previously, radiologists were writing reports with pens and watching stacks of films from the light box in a dark room. Patients came to the technician with handwritten prescriptions and registration information. Now, we have jumped into a new era. The workflow is greatly facilitated by automatic machine-facilitated workflow processes. Patients receive digital imaging. Radiologists sit in front of a computer and view images from picture archiving and communication sytem (PACS). Quantitative radiology and precision radiology have become the pursuit of us in our clinical practice. Data show that CT machine possession per million population in China increased from 5.5 in 2007 to 8.6 in 2010, although still low compared to the developed countries. Health-care PACS informational process market survey in 2015 showed that about 50% general public hospitals and 33% community hospitals have PACS. To improve the radiological service in China, multiple factors should be taken care of, not only the equipment, but also the education of radiologists/technicians, the facilitation and optimization of workflow, and the appropriateness and efficiency of patient care. With more than 13.8 billion population and varied economical situation in China, we are still facing great challenges. It is our goal to improve the quality of medical service, especially in rural areas. We will also encourage private health centres, imaging centres and Web-based consulting services in addition to established public health centres. We would like to stretch our arms towards globalization and welcome international communication and collaboration.

Interlude 1: Experience of Chinese traditional music
Multiparameteric analysis in imaging liver disease
F.-H. Yan; Shanghai/CN
Learning Objectives

1. To give an introduction to the imaging features of common liver diseases in China.
2. To understand the advantages of multiparameteric analysis of liver imaging.
3. To show some interesting cases.
4. To discuss the clinical application of multiparametric analysis in abdominal pathologies.


China has a serious chronic liver disease burden, especially HBV/HCV-related cirrhosis and hepatocellular carcinoma (HCC). According to the population-based cancer registration data of Chinese National Central Cancer Registry, HCC is the third most common cancer and the second most cause of cancer-related deaths. Over 80% of HCC cases develop in cirrhotic liver by a process of multistep hepatocarcinogenesis, from regenerative nodule (RN) , to dysplastic nodule (DN), and finally to overt HCC. Multiparametric imaging plays an important role for HCC surveillance, characterization with benign cirrhosis-associated nodules and noninvasively diagnostic algorithm in these extremely high-risk patients. To date, the diagnosis of HCC is mainly based on its haemodynamic hallmarks, i.e. arterial hyperenhancement followed by venous/delayed washout on multiphasic CT and MRI. It forms the foundation of current Western and Eastern HCC practice guidelines. Ancillary features, especially the presence of restricted diffusion on diffusion-weighted imaging (DWI), are preferred for early detection of HCC and to increase diagnostic confidence. MR imaging with hepatocyte-specific contrast agents has shown great promise in identifying high-grade DN and early HCC prior to neo-arterialization and progression to overt HCC. It is gradually changing the standard of diagnosis of HCC and may well be endorsed in future guidelines. Additionally, quantitative imaging techniques, including dual-energy CT, US- or MRI-based elastography, have demonstrated increasing roles for the evaluation of diffuse liver disease such as liver fibrosis, siderosis and nonalcoholic steatohepatitis.

Multimodality imaging for insulinoma detection
H.-D. Xue; Beijing/CN
Learning Objectives

1. To give a review of "insulinoma imaging" in the past.
2. To understand the advantages of perfusion CT, multiparametric MR and PET/CT with special tracers.
3. To discuss the potential clinical workflow and decision tree for imaging patients with suspected insulinomas and other functional pancreatic endocrine tumours.


Insulinomas are the most common hyper-functioning pancreatic endocrine tumours. Clinically, patients with insulinomas are characterized by hypoglycaemic symptoms caused by insulin overproduction. Insulinomas are generally benign and curable with surgery. Recurrence after resection is rare. Multiple lesions can be seen in patients with multiple endocrine neoplasia (MEN) syndrome. For patients with clinically suspected insulinomas, determining the location and number of tumours accurately with imaging workup allows for less invasive surgeries. Historically, detection of insulinomas by cross-sectional imaging has been difficult. Nowadays, a greater sensitivity for insulinoma detection has been reported with state-of-the-art CT and MR scanners. Routine dual-phase enhanced CT has relatively low sensitivity for insulinoma detection because of the presence of isoattenuating tumours, which account for around 25% of all tumours. Volume perfusion CT (VPCT) of the pancreas reveals increased blood flow of the tumour, as most ‘isoattenuating tumours’ show transient hyperenhancement, which may not be captured by traditional scanning protocols. Intra-individual comparative study shows that VPCT and mp-MR had similar high accuracy in insulinoma detection, and both were significantly more accurate than biphasic CECT. The intrinsic superior soft-tissue contrast with MRI helps to identify the tumour more easily without using contrast agents. Since insulinomas are quite small, the improved signal to noise of 3T MR compared to 1.5T and the introduction of high-resolution DWI are key factors for the improved diagnostic performance compared to the earlier reports.

Interventional therapy for hepatocellular carcinoma: Chinese experiences
X.-G. Li; Beijing/CN
Learning Objectives

1. To learn the epidemiology and special features of hepatocellular carcinoma in China.
2. To review the various interventional procedures and compare their efficacies in treating patients with hepatocellular carcinoma.
3. To understand the indications and clinical applications of interventional procedures to treat hepatocellular carcinoma.


In China, due to the high prevalence of HBV infection, the incidence rate of hepatocellular carcinoma is still very high. According to the 2017 Chinese cancer statistics report, HCC ranks 3rd and 6th in the incidence of cancer of men and women in urban areas, and it ranks 3rd in the mortality rate. The features of Chinese HCC patients include background of liver cirrhosis, relatively late stage, large tumor size and high rate of portal vein thrombus. Less than 20% of patients are candidates for liver resection or transplantation, so interventional treatment plays an important role in the management of HCC. Y90 microspheres are not available in the Mainland of China, so TARE is not routinely performed except for those in Hong Kong or Taiwan. In recent several years, DEB-TACE is accepted as an alternative for cTACE in selected patients. Thermal ablation including RFA, MWA and cryoablation is usually performed in combination with TACE. For lesions located at high-risk sites, PEI or 125I seed implantation is a good option and has achieved satisfactory results. Portal vein recanalization with 125I seed loading stent or endovascular RFA is attempted in some tertiary centres.

Interlude 2: Experience of Chinese tea culture
Application of artificial intelligence in prostate imaging
X.-Y. Wang; Beijing/CN
Learning Objectives

1. To learn about the basics and feasibility of applications of artificial intelligence.
2. To understand the clinical value of applying artificial intelligence in detection, prognosis and treatment evaluation of prostate cancer.
3. To learn the technique and clinical application of artificial intelligence in abdominal pelvic imaging.


A prostate mpMRI-CAD system has been developed. It can be used for identification of clinically significant prostate cancer with high prediction efficacy. It was validated by a multicentre clinical trial. After integrating the CAD system into the diagnostic process as a second reader, the performance of less experienced readers can be significantly improved.

Radiomics nomogram to predict lymph node metastasis in colorectal cancer
Z. Liu; Guangzhou/CN
Learning Objectives

1. To learn about the basics of radiomics and feasibility of clinical applications.
2. To understand the clinical value of radiomics nomogram in predicting lymph node metastasis in colorectal cancer.
3. To learn the potential applications of radiomics in various oncological studies.


To develop and validate a radiomics nomogram for preoperative predicting of lymph node (LN) metastasis in colorectal cancer (CRC). The prediction model was developed in a primary cohort consisting of 326 clinicopathologically confirmed CRC patients from January 2007 to April 2010. Radiomics features were extracted from portal venous-phase computed tomography (CT) of CRC. Lasso regression model was used for data dimension reduction, feature selection and radiomics signature building. Then, multivariable logistic regression analysis was used to develop the predicting model, with radiomics signature, CT-reported LN status and independent clinicopathological risk factors incorporated, which was presented with a radiomics nomogram. The performance of the nomogram was assessed with respect to its calibration, discrimination and clinical usefulness. Internal validation was assessed. An independent validation cohort contained 200 consecutive patients from May 2010 to December 2011. The radiomics signature consisting of 24 selected features was significantly associated with LN status (p<0.0001 for both the primary and validation cohorts). The predictors contained in the individualized prediction nomogram included the radiomics signature, CT-reported LN status and carcinoembryonic antigen (CEA) level. The addition of histological grade into the nomogram failed to show incremental prognostic value. The model showed good discrimination with a C-index of 0.736 (0.759 and 0.766 through internal validation), as well as a good calibration. Application of the nomogram in the validation cohort still gave good discrimination (C index=0.778 [95%CI: 0.769-0.787]) and good calibration. Decision curve analysis demonstrated that the radiomics nomogram had clinical usefulness. This study presents a radiomics nomogram incorporating the radiomics signature, CT-reported LN status and clinical risk factors, which can be conveniently used to facilitate the preoperative individualized prediction of LN metastasis in CRC.