This study aims to predict expression of Ki67 molecular marker in pancreatic cystic neoplasm using radiomics. We firstly manually segmented tumor area in multi-detector computed tomography (MDCT) images. Then 409 high-throughput features were automatically extracted and the least absolute shrinkage selection operator (LASSO) regression model was used for feature selection. After 200 bootstrapping repetitions of LASSO, 20 most frequently selected features made up the optimal feature set. Then 200 bootstrapping repetitions of support vector machine (SVM) classifier with 10-fold cross-validation were used to avoid overfitting and accurately predict the Ki67 expression. The highest prediction accuracy could achieve 85.29% and the highest area under the receiver operating characteristic curve (AUC) was 91.54% with a sensitivity (SENS) of 81.88% and a specificity (SPEC) of 86.75%. According to the results of experiment, the feasibility of predicting expression of Ki67 in pancreatic cystic neoplasm based on radiomics was verified.
ObjectiveTo investigate the guiding value of preoperative imaging and intraoperative rapid pathology in the diagnosis and treatment of pancreatic cystic neoplasm (PCN).MethodsThe clinical data of 205 patients with PCN diagnosed by pathology from July 14, 2003 to July 31, 2018 were analyzed retrospectively. The precise and fuzzy diagnostic rate and misdiagnosis rate of PCN by preoperative imaging and intraoperative rapid pathology were analyzed.ResultsThe most commonly used preoperative imaging methods were ultrasound and CT, in 146 cases (82.95%) and 141 cases (80.11%), respectively. There were 54 cases (30.68%) with MRI. Of them, 47 cases were examined by single examination, 129 cases received combined examination, of which 123 cases (95.35%) were examined by ultrasound combined with CT. The precise and fuzzy diagnostic rate of PCN by ultrasound, CT, and MRI were 81.51% (119/146), 81.56% (115/141), and 87.04% (47/54), respectively. Comparison of ultrasound with CT and MRI showed statistical significance (χ2=47.747, P<0.001; χ2=11.873, P=0.018), but no significant difference was observed between CT and MRI (χ2=5.012, P>0.05). In 27 cases of false diagnosis by ultrasound, no obvious abnormality was found in 14 cases (51.85%), followed by misdiagnosis as pancreatic pseudocyst (11 cases, 40.74%). Of the 26 cases misdiagnosed as pancreatic cancer by CT, 57.69% (15 cases) were misdiagnosed as pancreatic cancer; 7 cases were misdiagnosed by MRI, 42.86% (3 cases) of patients were misdiagnosed as pancreatic cancer and pancreatic pseudocyst. Thirty-one cases were misdiagnosed by intraoperative rapid pathology, and most of them misdiagnosed as pancreatic pseudocyst (10 cases, 32.26%). The next was SPN misdiagnosed as pancreatic neuroendocrine tumor (7 cases, 22.58%). The precise and fuzzy diagnostic rates of PCN were 81.58% (124/152), 86.84% (132/152), and 97.37% (148/152) in preoperative imaging, intraoperative rapid pathology, and preoperative imaging combined with intraoperative rapid pathology, while the misdiagnostic rates were 18.42% (28/152), 13.16% (20/152), and 2.63% (4/152), respectively.ConclusionsIn preoperative imaging and intraoperative rapid pathological examination, it is possible that ultrasound could not find PCN lesions. CT and MRI are most likely to be misdiagnosed as pancreatic cancer. Intraoperative rapid pathological examination misdiagnosed as pancreatic pseudocyst is most common. Perfect preoperative imaging and rapid intraoperative pathology can improve the correct diagnosis rate of PCN and avoid unreasonable surgical intervention measures.
ObjectiveTo summarize the experience of diagnosis and treatment of pancreatic cystic neoplasms (PCN).MethodThe general information, imaging findings, preoperative blood biochemical indexes, tumor markers, operation related data, postoperative complications, and pathological results of patients with PCN confirmed by pathological examination in Affiliated Hospital of Xuzhou Medical University from January 1, 2014 to December 31, 2019 were retrospectively analyzed.ResultsAmong the 158 patients who underwent surgical resection, the male to female ratio was 1∶1.6, and the median age was 56.2 years (range 26 to 82 years). Serous cystadenoma (SCN) was the most common, accounting for 34.2% (n=54), followed by mucinous cystadenoma (MCN) and intraductal papillary mucinous neoplasm (IPMN), accounting for 26.6% (n=42) and 20.2% (n=32). The proportion of solid pseudopapillary tumor (SPT) was the smallest, which was 19.0% (n=30). The PCN malignant rate was 19.0% (n=30), with the highest malignant rate being IPMN (43.8%, n=14), and the lowest being SCN (3.7%, n=2). Patients with PCN malignant group had significantly higher serum CA19-9, CEA, CA125, and neutrophil/lymphocyte ratio (NLR) than those in the benign group（all P<0.05）. The preoperative diagnosis of PCN was ambiguous, and the rate of misdiagnosis was high. The correct diagnosis rate of accurate preoperative specific classification was only 31.6% (n=50), and the lowest SCN (14.8%, n=8). High CA19-9, high NLR value, tumor location in the head/neck of pancreas and enhanced solid component were independent predictors of malignant PCN (P<0.05).ConclusionImproving the preoperative diagnosis ability of PCN and accurately determining the PCN classification before operation are helpful for the standardized treatment of PCN.
Objective To summarize the research progress in the diagnosis and treatment of pancreatic cystic neoplasms (PCNs). Method The guidelines and literatures related to the diagnosis and treatment of PCNs were collected and reviewed. Results At present, there was still no clear method to distinguish the types of PCNs and their benign and malignant, and there was still a dispute between domestic and foreign guidelines on the diagnosis and treatment of PCNs. Conclusion Clinical researchers still need to carry out more research, provide higher quality evidence, resolve the disputes existing in different guidelines, standardize the diagnosis and treatment process of PCNs, thus, PCNs can be identified early, diagnosed accurately and intervened in time.