Systematic approach for assessment of imaging features in chronic pancreatitis: a feasibility and validation study from the Scandinavian Baltic Pancreatic Club (SBPC) database

Abdominal Radiology(2020)

引用 10|浏览14
暂无评分
摘要
Purpose There is an unmet need for new systems with quantitative pancreatic imaging assessments to support better diagnosis and understand development of chronic pancreatitis (CP). The aims were to present such an approach for assessment of imaging features in CP, to apply this system in a multi-center cohort of CP patients (feasibility study), and to report inter-reader agreement between expert radiologists (validation study). Methods The feasibility study included pancreatic computed tomography (CT) or magnetic resonance imaging (MRI) from 496 patients with definitive CP in the Scandinavian Baltic Pancreatic Club (SBPC) database. Images were assessed according to the new SBPC imaging system (quantitative assessments of ductal and parenchymal features). Inter-reader agreement of reported imaging parameters was investigated for 80 CT and 80 MRI examinations by two expert radiologists. Results Reporting of the imaging features into the imaging system was deemed feasible for > 80% of CT and > 90% of MRI examinations. Quantitative assessments of main pancreatic duct diameters, presence/number/diameter of calcifications, and gland diameters had high levels of inter-reader agreement with κ -values of 0.75–0.87 and intraclass correlation coefficients of 0.74–0.97. The more subjective assessments, e.g., irregular main pancreatic duct and dilated side-ducts, had poor to moderate agreement with κ -values of 0.03–0.44. Conclusion The presented system provides a feasible mean for systematic assessment of CP imaging features. Imaging parameters based on quantitative assessment, as opposed to subjective assessments, have better reproducibility and should be preferred in the development of new grading systems for understanding pathophysiology and disease progression in CP.
更多
查看译文
关键词
Chronic pancreatitis, Imaging, Diagnosis, Severity, Validation
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要