Imaging features of primary hepatic sarcomatoid carcinoma: Differentiation from hepatocellular carcinoma and intrahepatic cholangiocarcinoma on CT: A preliminary study.

Purpose: Primary hepatic sarcomatoid carcinoma (PHSC) is a rare type of malignant tumor in the liver. Nevertheless, few studies have focused on the imaging diagnosis of PHSC. In this study, we collected clinical and computed tomography (CT) imaging data of PHSC from two institutions, aiming to investigate the clinical and radiological characteristics of PHSC. Methods: We retrospectively investigated the clinical characteristics and CT features of 22 PHSC patients (19 males and 3 females; mean age, 63.4 years; range, 49 to 76 years), 95 hepatocellular carcinoma (HCC) patients and 50 intrahepatic cholangiocarcinoma (ICC) patients. Two radiologists independently evaluated the CT features of the three groups. Subsequently, we analyzed the differences in the clinical characteristics and CT features between the PHSC and control groups. Results: Most PHSCs were larger than 5 cm (72.7%). PHSC mainly showed irregular (81.8%), heterogeneous (100%) masses with ill-defined (72.7%) borders with necrosis (86.4%) on CT, which are more common CT features versus HCC (p < 0.001). In the arterial phase, PHSC always showed noticeable heterogeneous enhancement (100.0%), mainly manifesting as partial arterial phase hyperenhancement (APHE) (86.4%). The enhancement patterns of PHSC mainly included delayed progressive enhancement (72.7%), nonperipheral washout (22.7%), and unclassified enhancement (4.5%), which were significantly different from the HCC enhancement pattern but similar to the enhancement pattern of ICC. In addition, vein tumor thrombus (18.2%), intrahepatic metastasis (27.3%), and lymphadenopathy (27.3%) were relatively common in PHSC. Furthermore, most PHSC tumors classified as LR-M (66.7%) were similar to ICC. Conclusions: PHSC generally presents as irregularly large masses with necrosis, intrahepatic metastasis, and lymphadenopathy. The CT enhancement of PHSC is mainly part of APHE and delayed progressive enhancement.

Classification prediction of pancreatic cystic neoplasms based on radiomics deep learning models.

BACKGROUND: Preoperative prediction of pancreatic cystic neoplasm (PCN) differentiation has significant value for the implementation of personalized diagnosis and treatment plans. This study aimed to build radiomics deep learning (DL) models using computed tomography (CT) data for the preoperative differential diagnosis of common cystic tumors of the pancreas. METHODS: Clinical and CT data of 193 patients with PCN were collected for this study. Among these patients, 99 were pathologically diagnosed with pancreatic serous cystadenoma (SCA), 55 were diagnosed with mucinous cystadenoma (MCA) and 39 were diagnosed with intraductal papillary mucinous neoplasm (IPMN). The regions of interest (ROIs) were obtained based on manual image segmentation of CT slices. The radiomics and radiomics-DL models were constructed using support vector machines (SVMs). Moreover, based on the fusion of clinical and radiological features, the best combined feature set was obtained according to the Akaike information criterion (AIC) analysis. Then the fused model was constructed using logistic regression. RESULTS: For the SCA differential diagnosis, the fused model performed the best and obtained an average area under the curve (AUC) of 0.916. It had a best feature set including position, polycystic features (≥6), cystic wall calcification, pancreatic duct dilatation and radiomics-DL score. For the MCA and IPMN differential diagnosis, the fused model with AUC of 0.973 had a best feature set including age, communication with the pancreatic duct and radiomics score. CONCLUSIONS: The radiomics, radiomics-DL and fused models based on CT images have a favorable differential diagnostic performance for SCA, MCA and IPMN. These findings may be beneficial for the exploration of individualized management strategies.

Deep Learning Radiomics to Predict Regional Lymph Node Staging for Hilar Cholangiocarcinoma.

BACKGROUND: Our aim was to establish a deep learning radiomics method to preoperatively evaluate regional lymph node (LN) staging for hilar cholangiocarcinoma (HC) patients. METHODS AND MATERIALS: Of the 179 enrolled HC patients, 90 were pathologically diagnosed with lymph node metastasis. Quantitative radiomic features and deep learning features were extracted. An LN metastasis status classifier was developed through integrating support vector machine, high-performance deep learning radiomics signature, and three clinical characteristics. An LN metastasis stratification classifier (N1 vs. N2) was also proposed with subgroup analysis. RESULTS: The average areas under the receiver operating characteristic curve (AUCs) of the LN metastasis status classifier reached 0.866 in the training cohort and 0.870 in the external test cohorts. Meanwhile, the LN metastasis stratification classifier performed well in predicting the risk of LN metastasis, with an average AUC of 0.946. CONCLUSIONS: Two classifiers derived from computed tomography images performed well in predicting LN staging in HC and will be reliable evaluation tools to improve decision-making.

Differentiating Hepatic Epithelioid Angiomyolipoma From Hepatocellular Carcinoma and Focal Nodular Hyperplasia via Radiomics Models.

Background: We conduct a study in developing and validating two radiomics-based models to preoperatively distinguish hepatic epithelioid angiomyolipoma (HEAML) from hepatic carcinoma (HCC) as well as focal nodular hyperplasia (FNH). Methods: Totally, preoperative contrast-enhanced computed tomography (CT) data of 170 patients and preoperative contrast-enhanced magnetic resonance imaging (MRI) data of 137 patients were enrolled in this study. Quantitative texture features and wavelet features were extracted from the regions of interest (ROIs) of each patient imaging data. Then two radiomics signatures were constructed based on CT and MRI radiomics features, respectively, using the random forest (RF) algorithm. By integrating radiomics signatures with clinical characteristics, two radiomics-based fusion models were established through multivariate linear regression and 10-fold cross-validation. Finally, two diagnostic nomograms were built to facilitate the clinical application of the fusion models. Results: The radiomics signatures based on the RF algorithm achieved the optimal predictive performance in both CT and MRI data. The area under the receiver operating characteristic curves (AUCs) reached 0.996, 0.879, 0.999, and 0.925 for the training as well as test cohort from CT and MRI data, respectively. Then, two fusion models simultaneously integrated clinical characteristics achieved average AUCs of 0.966 (CT data) and 0.971 (MRI data) with 10-fold cross-validation. Through decision curve analysis, the fusion models were proved to be excellent models to distinguish HEAML from HCC and FNH in comparison between the clinical models and radiomics signatures. Conclusions: Two radiomics-based models derived from CT and MRI images, respectively, performed well in distinguishing HEAML from HCC and FNH and might be potential diagnostic tools to formulate individualized treatment strategies.

Preoperative Prediction Power of Imaging Methods for Microvascular Invasion in Hepatocellular Carcinoma: A Systemic Review and Meta-Analysis.

Background: To compare the predictive power between radiomics and non-radiomics (conventional imaging and functional imaging methods) for preoperative evaluation of microvascular invasion (MVI) in hepatocellular carcinoma (HCC). Methods: Comprehensive publications were screened in PubMed, Embase, and Cochrane Library. Studies focusing on the discrimination values of imaging methods, including radiomics and non-radiomics methods, for MVI evaluation were included in our meta-analysis. Results: Thirty-three imaging studies with 5,462 cases, focusing on preoperative evaluation of MVI status in HCC, were included. The sensitivity and specificity of MVI prediction in HCC were 0.78 [95% confidence interval (CI): 0.75-0.80; I 2 = 70.7%] and 0.78 (95% CI: 0.76-0.81; I 2 = 0.0%) for radiomics, respectively, and were 0.73 (95% CI: 0.71-0.75; I 2 = 83.7%) and 0.82 (95% CI: 0.80-0.83; I 2 = 86.5%) for non-radiomics, respectively. The areas under the receiver operation curves for radiomics and non-radiomics to predict MVI status in HCC were 0.8550 and 0.8601, respectively, showing no significant difference. Conclusion: The imaging method is feasible to predict the MVI state of HCC. Radiomics method based on medical image data is a promising application in clinical practice and can provide quantifiable image features. With the help of these features, highly consistent prediction performance will be achieved in anticipation.

Contrast-enhanced CT radiomics for preoperative evaluation of microvascular invasion in hepatocellular carcinoma: A two-center study.

BACKGROUND: The present study constructed and validated the use of contrast-enhanced computed tomography (CT)-based radiomics to preoperatively predict microvascular invasion (MVI) status (positive vs negative) and risk (low vs high) in patients with hepatocellular carcinoma (HCC). METHODS: We enrolled 637 patients from two independent institutions. Patients from Institution I were randomly divided into a training cohort of 451 patients and a test cohort of 111 patients. Patients from Institution II served as an independent validation set. The LASSO algorithm was used for the selection of 798 radiomics features. Two classifiers for predicting MVI status and MVI risk were developed using multivariable logistic regression. We also performed a survival analysis to investigate the potentially prognostic value of the proposed MVI classifiers. RESULTS: The developed radiomics signature predicted MVI status with an area under the receiver operating characteristic curve (AUC) of .780, .776, and .743 in the training, test, and independent validation cohorts, respectively. The final MVI status classifier that integrated two clinical factors (age and α-fetoprotein level) achieved AUC of .806, .803, and .796 in the training, test, and independent validation cohorts, respectively. For MVI risk stratification, the AUCs of the radiomics signature were .746, .664, and .700 in the training, test, and independent validation cohorts, respectively, and the AUCs of the final MVI risk classifier-integrated clinical stage were .783, .778, and .740, respectively. Survival analysis showed that our MVI status classifier significantly stratified patients for short overall survival or early tumor recurrence. CONCLUSIONS: Our CT radiomics-based models were able to predict MVI status and MVI risk of HCC and might serve as a reliable preoperative evaluation tool.