A Transformer-Based microvascular invasion classifier enhances prognostic stratification in HCC following radiofrequency ablation.

BACKGROUND & AIMS: We aimed to develop a Transformer-based deep learning (DL) network for prognostic stratification in hepatocellular carcinoma (HCC) patients undergoing RFA. METHODS: A Swin Transformer DL network was trained to establish associations between magnetic resonance imaging (MRI) datasets and the ground truth of microvascular invasion (MVI) based on 696 surgical resection (SR) patients with solitary HCC ≤3 cm, and was validated in an external cohort (n = 180). The multiphase MRI-based DL risk outputs using an optimal threshold of .5 was employed as a MVI classifier for prognosis stratification in the RFA cohort (n = 180). RESULTS: Over 90% of all enrolled patients exhibited hepatitis B virus infection. Liver cirrhosis was significantly more prevalent in the RFA cohort compared to the SR cohort (72.2% vs. 44.1%, p < .001). The MVI risk outputs exhibited good performance (area under the curve values = .938 and .883) for predicting MVI in the training and validation cohort, respectively. The RFA patients at high risk of MVI classified by the MVI classifier demonstrated significantly lower recurrence-free survival (RFS) and overall survival rates at 1, 3 and 5 years compared to those classified as low risk (p < .001). Multivariate cox regression modelling of a-fetoprotein > 20 ng/mL [hazard ratio (HR) = 1.53; 95% confidence interval (95% CI): 1.02-2.33, p = .047], high risk of MVI (HR = 3.76; 95% CI: 2.40-5.88, p < .001) and unfavourable tumour location (HR = 2.15; 95% CI: 1.40-3.29, p = .001) yielded a c-index of .731 (bootstrapped 95% CI: .667-.778) for evaluating RFS after RFA. Among the three risk factors, MVI was the most powerful predictor for intrahepatic distance recurrence. CONCLUSIONS: The proposed MVI classifier can serve as a valuable imaging biomarker for prognostic stratification in early-stage HCC patients undergoing RFA.

Diagnostic performance of gadoxetic acid-enhanced abbreviated magnetic resonance imaging protocol in small hepatocellular carcinoma (≤2 cm) in high-risk patients.

BACKGROUND: Biannual Ultrasound showed insufficient sensitivity in detecting small or early-stage hepatocellular carcinoma (HCC). Abbreviated magnetic resonance imaging (A-MRI) protocols with fewer sequences demonstrated higher HCC detection sensitivity than ultrasound with acceptable cost and examination time. PURPOSE: To compare the diagnostic performance of gadoxetic acid-enhanced A-MRI with a full sequence MRI (F-MRI) protocol for small HCC (≤2 cm) in cirrhotic or hepatitis B virus-infected high-risk patients. MATERIAL AND METHODS: Two hundred and four consecutive patients with 166 pathologically confirmed small HCC who underwent preoperative gadoxetic acid-enhanced MRI were retrospectively included. A-MRI set comprised T1-weighted hepatobiliary phase imaging, T2-weighted imaging, diffusion-weighted imaging and apparent diffusion coefficient mapping. Two independent radiologists blinded to clinical data assessed the A-MRI set and F-MRI set. Per-patient HCC and per-lesion HCC diagnostic performance were compared. RESULTS: Per-patient HCC detection sensitivity of A-MRI set was 93.8% and 91.2% for observer 1 and observer 2, and, for the F-MRI set, the per-patient HCC detection sensitivity was 96.6% and 95.2%, respectively. There was no significant difference in per-patient sensitivity, specificity and per-lesion HCC detection sensitivity between the two imaging sets for both readers. (P = 0.06-0.25) The A-MRI set showed higher sensitivity on HCC without arterial phase hyperenhancement, and the F-MRI set demonstrated with higher sensitivity on HCC with arterial phase hyperenhancement (P < 0.05). CONCLUSION: A-MRI using diagnostic criteria including hypointensity on hepatobiliary phase plus mild to moderate hyperintensity on T2-weighted imaging or restricted diffusion demonstrated comparable sensitivity and specificity for small HCC compared to the F-MRI protocol in high-risk patients.

A Multiparametric Fusion Deep Learning Model Based on DCE-MRI for Preoperative Prediction of Microvascular Invasion in Intrahepatic Cholangiocarcinoma.

BACKGROUND: Assessment of microvascular invasion (MVI) in intrahepatic cholangiocarcinoma (ICC) by using a noninvasive method is an unresolved issue. Deep learning (DL) methods based on multiparametric fusion of MR images have the potential of preoperative assessment of MVI. PURPOSE: To investigate whether a multiparametric fusion DL model based on MR images can be used for preoperative assessment of MVI in ICC. STUDY TYPE: Retrospective. POPULATION: A total of 519 patients (200 females and 319 males) with a single ICC were categorized as a training (n = 361), validation (n = 90), and an external test cohort (n = 68). FIELD STRENGTH/SEQUENCE: A 1.5 T and 3.0 T; axial T2-weighted turbo spin-echo sequence, diffusion-weighted imaging with a single-shot spin-echo planar sequence, and dynamic contrast-enhanced (DCE) imaging with T1-weighted three-dimensional quick spoiled gradient echo sequence. ASSESSMENT: DL models of multiparametric fusion convolutional neural network (CNN) and late fusion CNN were both constructed for evaluating MVI in ICC. Gradient-weighted class activation mapping was used for visual interpretation of MVI status in ICC. STATISTICAL TESTS: The DL model performance was assessed through the receiver operating characteristic curve (ROC) analysis, and the area under the ROC curve (AUC) with the accuracy, sensitivity, and specificity were measured. P value < 0.05 was considered as statistical significance. RESULTS: In the external test cohort, the proposed multiparametric fusion DL model achieved an AUC of 0.888 with an accuracy of 86.8%, sensitivity of 85.7%, and specificity of 87.0% for evaluating MVI in ICC, and the positive predictive value and negative predictive value were 63.2% and 95.9%, respectively. The late fusion DL model achieved a lower AUC of 0.866, with an accuracy of 83.8%, sensitivity of 78.6%, specificity of 85.2% for evaluating MVI in ICC. DATA CONCLUSION: Our DL model based on multiparametric fusion of MRI achieved a good diagnostic performance in the evaluation of MVI in ICC. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 2.

Microvascular invasion of small hepatocellular carcinoma can be preoperatively predicted by the 3D quantification of MRI.

OBJECTIVES: To explore the importance of three-dimensional (3D) quantitative analysis during gadoxetic acid-enhanced magnetic resonance imaging (MRI) of microvascular invasion (MVI) and early recurrence (< 2 years) after surgery of single hepatocellular carcinoma (HCC) ≤ 3 cm. METHODS: Two hundred fourteen patients with pathologically confirmed HCC (training cohort: n = 169; validation cohort: n = 45) were included retrospectively. The 3D quantitative parameters (volume, sphericity, and compacity) and conventional MRI features were analyzed. The significant predictors for MVI were identified using univariate and multivariate logistic regression analyses. Nomograms were constructed from the prediction model, and the relationship between the significant predictors and early recurrence rates was evaluated using the Kaplan-Meier method. RESULTS: Tumor sphericity (odds ratio [OR] = 0.000; p < 0.001), non-smooth tumor margin (OR = 3.353; p = 0.015), and peritumoral hypointensity on hepatobiliary phase (HBP) (OR = 14.067; p = 0.003) were independent significant factors for MVI. When these three factors were combined, the diagnostic specificity of the training and validation cohorts was 97.0 (128/132) and 87.9 (29/33), respectively. The nomogram based on the predictive model performed satisfactorily in the training (C-index: 0.885) and validation (C-index: 0.869) cohorts. Early recurrence rates of patients with two or three significant factors were significantly higher than those with none in the training (29.1% vs. 10.2%, p = 0.007) and validation (36.4% vs. 6.7%, p = 0.037) cohorts. CONCLUSIONS: Lower sphericity combined with non-smooth tumor margin and peritumoral hypointensity on HBP are potential predictive factors for MVI and associated with early recurrence after surgery of HCC ≤ 3 cm. KEY POINTS: • Lower sphericity, non-smooth tumor margin, and peritumoral hypointensity on HBP were important indicators of the occurrence of MVI in HCC. • The combinational model prepared from these findings satisfactorily predicted MVI, and the presence of these predictors was associated with an early recurrence rate after surgical resection in HCC patients. • This model could help clinicians in the preoperative management of small HCC ≤ 3 cm.

Hepatobiliary phase images of gadoxetic acid-enhanced MRI may improve accuracy of predicting the size of hepatocellular carcinoma at pathology.

BACKGROUND: Gadoxetic acid-enhanced magnetic resonance imaging (MRI) has been widely used in clinical practice. However, scientific evidence is lacking for recommending a particular sequence for measuring tumor size. PURPOSE: To retrospectively compare the size of hepatocellular carcinoma (HCC) measured on different gadoxetic acid-enhanced MRI sequences using pathology as a reference. MATERIAL AND METHODS: A total of 217 patients with single HCC who underwent gadoxetic acid-enhanced MRI before surgery were included. The size of the HCC was measured by two abdominal radiologists independently on the following sequences: T1-weighted; T2-weighted; b-500 diffusion-weighted imaging (DWI); and arterial, portal venous, transitional, and hepatobiliary phases. Tumor size measured on MRI was compared with pathological size by using Pearson correlation coefficient, independent-sample t test, and Bland-Altman plot. Agreement between two readers was evaluated with intraclass correlation coefficient (ICC). RESULTS: Correlation between the MR images and pathology was high for both readers (0.899-0.955). Absolute error between MRI and pathologic assessment was lowest on hepatobiliary phase images for both readers (reader 1, 2.8±4.2 mm; reader 2, 3.2±3.4 mm) and highest on arterial phase images for reader 1 (4.9±4.4 mm) and DWI phase images for reader 2 (5.1±4.9 mm). Absolute errors were significantly different for hepatobiliary phase compared with other sequences for both readers (reader 1, P≤0.012; reader 2, P≤0.037). Inter-reader agreements for all sequence measurements were strong (0.971-0.997). CONCLUSION: The performance of gadoxetic acid-enhanced MRI sequences varied with HCC size, and the hepatobiliary phase may be optimal among these sequences.

Ultrashort time-to-echo T2* and T2* relaxometry for evaluation of lumbar disc degeneration: a comparative study.

BACKGROUND: To compare potential of ultrashort time-to-echo (UTE) T2* mapping and T2* values from T2*-weighted imaging for assessing lumbar intervertebral disc degeneration (IVDD),with Pfirrmann grading as a reference standard. METHODS: UTE-T2* and T2* values of 366 lumbar discs (L1/2-L5/S1) in 76 subjects were measured in 3 segmented regions: anterior annulus fibrosus, nucleus pulposus (NP), and posterior annulus fibrosus. Lumbar intervertebral discs were divided into 3 categories based on 5-level Pfirrmann grading: normal (Pfirrmann grade I),early disc degeneration (Pfirrmann grades II-III), and advanced disc degeneration (Pfirrmann grades IV-V). Regional differences between UTE-T2* and T2* relaxometry and correlation with degeneration were statistically analyzed. RESULTS: UTE-T2* and T2*value correlated negatively with Pfirrmann grades (P < 0.001). In NP, correlations with Pfirrmann grade were high with UTE-T2* values (r = - 0.733; P < 0.001) and moderate with T2* values (r = -0.654; P < 0.001). Diagnostic accuracy of detecting early IVDD was better with UTE-T2* mapping than T2* mapping (P < 0.05),with receiver operating characteristic analysis area under the curve of 0.715-0.876. CONCLUSIONS: UTE-T2* relaxometry provides another promising magnetic resonance imaging sequence for quantitatively evaluate lumbar IVDD and was more accurate than T2*mapping in the earlier stage degenerative process.

A predictive model integrating deep and radiomics features based on gadobenate dimeglumine-enhanced MRI for postoperative early recurrence of hepatocellular carcinoma.

PURPOSE: Hepatocellular carcinoma (HCC) is the most common liver cancer worldwide, and early recurrence of HCC after curative hepatic resection is indicative of poor prognoses. We aim to develop a predictive model for postoperative early recurrence of HCC based on deep and radiomics features from multi-phasic magnetic resonance imaging (MRI). MATERIALS AND METHODS: A total of 472 HCC patients were included and divided into the training (n = 378) and validation (n = 94) cohorts in the retrospective study. We separately extracted radiomics features and deep features from eight phases of gadoxetic acid-enhanced MRI and utilized the least absolute shrinkage and selection operator logistic regression algorithm for feature selection and model construction. We integrated the selected two types of features into a combined model and established a radiomics model as well as a deep learning (DL) model for comparison. RESULTS: In the training and validation cohorts, the combined model demonstrated better performance for stratifying patients at high risk of early recurrence (AUC of 0.911 and 0.840, accuracy of 0.779 and 0.777, sensitivity of 0.927 and 0.769, specificity 0.720 and 0.779) than the radiomics model (AUC of 0.740 and 0.780) and the DL model (AUC of 0.887 and 0.813). CONCLUSION: The combined model integrating deep and radiomics features from multi-phasic MRI is efficient for noninvasively stratifying patients at high risk of early HCC recurrence after resection.

Identifying response in colorectal liver metastases treated with bevacizumab: development of RECIST by combining contrast-enhanced and diffusion-weighted MRI.

OBJECTIVES: Response evaluation criteria in solid tumors (RECIST) often fail to identify clinically meaningful response to bevacizumab-containing therapy in colorectal liver metastasis (CRLM). This study aimed to develop RECIST by combining contrast-enhanced and diffusion-weighted magnetic resonance imaging (MRI). METHODS: A total of 126 patients with CRLM who underwent hepatic resection after bevacizumab-containing chemotherapy were split into initial analyses cohort (N = 42, with 76 indexed liver metastases) and validation cohort (N = 84). In lesion-based analyses, percentage decrease of arterial enhancement area and percentage increase of apparent diffusion coefficient (ADC) value from baseline to post-chemotherapy were measured. Their optimal cutoff values for distinguishing pathology-confirmed major and minor response were determined. Then, the developed RECIST (D-RECIST) was established by combining functional and size-based items. Survival relevance of D-RECIST and RECIST was examined in the validation cohort. RESULTS: Percentage decrease of arterial enhancement area and increase of ADC value significantly differed between lesions of pathologic major or minor response, with optimal cutoffs of approximately 33% and 19%, respectively. Patients defined as responders by D-RECIST had a significantly longer median disease-free survival (DFS) than non-responders (p = 0.021; 12.9 versus 8.6 months). No significant difference was observed with RECIST (p = 0.524). In a Cox regression model, D-RECIST- but not RECIST-defined responses independently predicted the DFS (p = 0.034 and 0.811). CONCLUSIONS: D-RECIST-defined responses provided significant prognostic information, and thus may serve as a better response evaluation approach than RECIST in CRLM treated with bevacizumab-containing therapy. KEY POINTS: • Changes in arterial enhancement area and apparent diffusion coefficient value are associated with pathological response in colorectal liver metastases treated with bevacizumab. • The MRI-based response criteria developed by combining size-based and functional features can provide significant prognostic information.

Multi-scale and multi-parametric radiomics of gadoxetate disodium-enhanced MRI predicts microvascular invasion and outcome in patients with solitary hepatocellular carcinoma ≤ 5 cm.

OBJECTIVES: To develop radiomics-based nomograms for preoperative microvascular invasion (MVI) and recurrence-free survival (RFS) prediction in patients with solitary hepatocellular carcinoma (HCC) ≤ 5 cm. METHODS: Between March 2012 and September 2019, 356 patients with pathologically confirmed solitary HCC ≤ 5 cm who underwent preoperative gadoxetate disodium-enhanced MRI were retrospectively enrolled. MVI was graded as M0, M1, or M2 according to the number and distribution of invaded vessels. Radiomics features were extracted from DWI, arterial, portal venous, and hepatobiliary phase images in regions of the entire tumor, peritumoral area ≤ 10 mm, and randomly selected liver tissue. Multivariate analysis identified the independent predictors for MVI and RFS, with nomogram visualized the ultimately predictive models. RESULTS: Elevated alpha-fetoprotein, total bilirubin and radiomics values, peritumoral enhancement, and incomplete or absent capsule enhancement were independent risk factors for MVI. The AUCs of MVI nomogram reached 0.920 (95% CI: 0.861-0.979) using random forest and 0.879 (95% CI: 0.820-0.938) using logistic regression analysis in validation cohort (n = 106). With the 5-year RFS rate of 68.4%, the median RFS of MVI-positive (M2 and M1) and MVI-negative (M0) patients were 30.5 (11.9 and 40.9) and > 96.9 months (p < 0.001), respectively. Age, histologic MVI, alkaline phosphatase, and alanine aminotransferase independently predicted recurrence, yielding AUC of 0.654 (95% CI: 0.538-0.769, n = 99) in RFS validation cohort. Instead of histologic MVI, the preoperatively predicted MVI by MVI nomogram using random forest achieved comparable accuracy in MVI stratification and RFS prediction. CONCLUSIONS: Preoperative radiomics-based nomogram using random forest is a potential biomarker of MVI and RFS prediction for solitary HCC ≤ 5 cm. KEY POINTS: • The radiomics score was the predominant independent predictor of MVI which was the primary independent risk factor for postoperative recurrence. • The radiomics-based nomogram using either random forest or logistic regression analysis has obtained the best preoperative prediction of MVI in HCC patients so far. • As an excellent substitute for the invasive histologic MVI, the preoperatively predicted MVI by MVI nomogram using random forest (MVI-RF) achieved comparable accuracy in MVI stratification and outcome, reinforcing the radiologic understanding of HCC angioinvasion and progression.

A magnetic resonance imaging modality for non-invasively distinguishing progression of liver fibrosis by visualizing hepatic platelet-derived growth factor receptor-beta expression in mice.

BACKGROUND AND AIM: Activated hepatic stellate cells (HSCs) are the most critical cells responsible for liver fibrosis, and platelet-derived growth factor (PDGF) is the most prominent mitogen for HSCs in fibrogenesis. This study aimed to explore the potential of gadolinium (Gd)-labeled cyclic peptides (pPB) targeting PDGF receptor-ß (PDGFR-ß) as a magnetic resonance imaging (MRI) radiotracer to identify the progression of liver fibrosis by imaging hepatic PDGFR-ß expression. METHODS: Mice treated with carbon tetrachloride (CCl4 ) were used to mimic hepatic fibrosis in vivo. The binding activity of FITC-labeled pPB to PDGFR-ß was assessed in cultured human HSCs (HSC-LX2). MRI was performed to visualize hepatic PDGFR-ß expression in mice with different degrees of liver fibrosis after Gd-labeled pPB was injected. RESULTS: Hepatic PDGFR-ß expression level was correlated with the severity of liver fibrosis, and the majority of cells expressing PDGFR-ß were found to be activated HSCs in fibrotic livers. Culture-activated human HSCs expressed abundant PDGFR-ß, and FITC-labeled pPB could bind to these cells in a concentration-dependent and time-dependent manner. With Gd-labeled pPB as a tracer, an MRI modality demonstrated that the relative hepatic T1-weighted MRI signal value progressively increased with the severity of hepatic fibrosis and reduced with remission. CONCLUSIONS: Hepatic PDGFR-ß expression reflects the progression of hepatic fibrosis, and MRI using Gd-labeled pPB as a tracer exhibits potential for distinguishing liver fibrosis staging in mice.

Recurrence After Curative Resection of Hepatitis B Virus-Related Hepatocellular Carcinoma: Diagnostic Algorithms on Gadoxetic Acid-Enhanced Magnetic Resonance Imaging.

Small recurrent hepatocellular carcinoma (HCC) can show atypical imaging patterns, and a specific diagnostic algorithm for HCC is lacking. This study aimed to better characterize postoperative recurrent HCCs <20 mm in size with gadoxetic acid-enhanced magnetic resonance imaging (MRI). We evaluated 373 newly developed nodules after hepatectomy in 204 HCC patients with chronic hepatitis B virus infection. The diagnostic performance of Liver Imaging Reporting and Data System (LI-RADS) version 2018 was calculated with gadoxetic acid-enhanced MRI to characterize recurrent HCC. Modified diagnostic algorithms were proposed by combining significant imaging biomarkers related to subcentimeter and 10-19 mm recurrence, and the algorithms were then compared with the LI-RADS system. A total of 256 recurrent HCCs (108 recurrent HCCs <10 mm in size; 148 recurrent HCCs 10-19 mm in size) were confirmed via histology or follow-up imaging. Nonrim arterial phase hyperenhancement (APHE) and 3 LI-RADS ancillary features (AFs; hepatobiliary phase hypointensity, mild-moderate T2 hyperintensity, and restricted diffusion) were significantly related to recurrent HCCs <20 mm in size according to a multivariate analysis. For subcentimeter recurrence, combining at least 2 of the 3 AFs only achieved better specificity (sensitivity, 83.3%; specificity, 87.7%) than the LR-4 category (sensitivity, 88.9%, P = 0.21; specificity, 70.8%, P = 0.006). For 10-19 mm recurrences, combining nonrim APHE and at least 1 of the 3 AFs achieved only a significantly enhanced sensitivity of 85.1% but a lower specificity of 86.5% compared with the LR-5 category (sensitivity: 63.5%, P < 0.001; specificity: 94.2%, P = 0.13). In conclusion, the diagnostic algorithms for subcentimeter and 10-19 mm recurrent HCCs should be stratified. Combining at least 2 AFs demonstrated comparable sensitivity with significantly enhanced specificity compared with the LR-4 category for characterizing subcentimeter recurrence.

A radiomics-based biomarker for cytokeratin 19 status of hepatocellular carcinoma with gadoxetic acid-enhanced MRI.

OBJECTIVES: We aimed to develop a radiomics-based model derived from gadoxetic acid-enhanced MR images to preoperatively identify cytokeratin (CK) 19 status of hepatocellular carcinoma (HCC). METHODS: A cohort of 227 patients with single HCC was classified into a training set (n = 159) and a time-independent validated set (n = 68). A total of 647 radiomic features were extracted from multi-sequence MR images. The least absolute shrinkage and selection operator regression and decision tree methods were utilized for feature selection and radiomics signature construction. A multivariable logistic regression model incorporating clinico-radiological features and the fusion radiomics signature was built for prediction of CK19 status by evaluating area under curve (AUC). RESULTS: In the whole cohort, 57 patients were CK19 positive and 170 patients were CK19 negative. By combining 11 and 6 radiomic features extracted in arterial phase and hepatobiliary phase images, respectively, a fusion radiomics signature achieved AUCs of 0.951 and 0.822 in training and validation datasets. The final combined model integrated a-fetoprotein levels, arterial rim enhancement pattern, irregular tumor margin, and the fusion radiomics signature, with a sensitivity of 0.818 and specificity of 0.974 in the training cohort and that of 0.769 and 0.818 in the validated cohort. The nomogram based on the combined model showed satisfactory prediction performance in training (C-index 0.959) and validation (C-index 0.846) dataset. CONCLUSIONS: The combined model based on a fusion radiomics signature derived from arterial and hepatobiliary phase images of gadoxetic acid-enhanced MRI can be a reliable biomarker for CK19 status of HCC. KEY POINTS: • Arterial rim enhancement pattern and irregular tumor margin on hepatobiliary phase on gadoxetic acid-enhanced MRI can be useful for evaluating CK19 status of HCC. • A radiomics-based model performed better than the clinico-radiological model both in training and validation datasets for predicting CK19 status of HCC. • The nomogram based on the fusion radiomics signature can be easily used for CK19 stratification of HCC.

Pre-TACE kurtosis of ADCtotal derived from histogram analysis for diffusion-weighted imaging is the best independent predictor of prognosis in hepatocellular carcinoma.

PURPOSE: To determine the feasibility of pre-TACE IVIM imaging based on histogram analysis for predicting prognosis in the treatment of unresectable hepatocellular carcinoma (HCC). MATERIALS AND METHODS: Fifty-five patients prospectively underwent 1.5T MRI 1 week before TACE. Histogram metrics for IVIM parameters and ADCs maps between responders and non-responders with mRECIST assessment were compared. Kaplan-Meier, log-rank tests and Cox proportional hazard regression model were used to correlate variables with time to progression (TTP). RESULTS: Mean (p = 0.022), median (p = 0.043), and 25th percentile (p < 0.001) of perfusion fraction (PF), mean (p < 0.001), median (p < 0.001), 25th percentile (p < 0.001) and 75th percentile (p = 0.001) of ADC(0,500), mean (p = 0.005), median (p = 0.008) and 25th percentile (p = 0.039) of ADCtotal were higher, while skewness and kurtosis of PF (p = 0.001, p = 0.005, respectively), kurtosis of ADC(0,500) and ADCtotal (p = 0.005, p = 0.001, respectively) were lower in responders compared to non-responders. Multivariable analysis demonstrated that mRECIST was associated with TTP independently, and kurtosis of ADCtotal had the best predictive performance for disease progression. CONCLUSION: Pre-TACE kurtosis of ADCtotal is the best independent predictor for TTP. KEY POINTS: • mRECIST was associated with TTP independently. • Lower kurtosis and higher mean for ADCs tend to have good response. • Pre-TACE kurtosis of ADC total is the best independent predictor for TTP.

Difference analysis in prevalence of incidental pancreatic cystic lesions between computed tomography and magnetic resonance imaging.

BACKGROUND: The purpose was to investigate the difference of detection rate of incidental pancreatic cystic lesions (PCLs) with computed tomography (CT) and magnetic resonance imaging (MRI) and to compare the difference between CT and MRI and to explore the effect of this difference on surgical resection. METHODS: We reviewed the diagnostic reports for incidental PCLs between 2013 and 2016. Images of PCLs would be re-evaluated. Clinical and imaging data were recorded. The chi-square and independent t-test were conducted for categorical and continuous variables. RESULTS: The prevalence of PCLs was 1.91% (1038/54210) and 3.36% (1282/38099) on CT and MRI respectively, and increased with increasing age (P < 0.001). No significant differences were found in the annual prevalence of PCLs on CT (P = 0.796) and MRI (P = 0.213) from 2013 to 2016 while the number of examinations was increasing every year. The annual detection rate of MRI for small PCLs (< 20 mm) was significantly higher than CT (P < 0.001), but was not significantly different for large PCLs (≥20 mm). The rate of surgical resection of PCLs (≥20 mm) in MRI group was higher than CT (55.2% vs. 37.0%, P < 0.001). CONCLUSIONS: The detection rate of PCLs on CT and MRI tended to be stable despite increasing scan volumes. Female had a slightly more frequency of PCLs than male. MRI detected more small PCLs(< 20 mm) and had higher impact on surgical resection of large PCL(≥20 mm) compared with CT.

Assessment of Microvascular Invasion of Hepatocellular Carcinoma with Diffusion Kurtosis Imaging.

Purpose To evaluate the potential role of diffusion kurtosis imaging and conventional magnetic resonance (MR) imaging findings including standard monoexponential model of diffusion-weighted imaging and morphologic features for preoperative prediction of microvascular invasion (MVI) of hepatocellular carcinoma (HCC). Materials and Methods Institutional review board approval and written informed consent were obtained. Between September 2015 and November 2016, 84 patients (median age, 54 years; range, 29-79 years) with 92 histopathologically confirmed HCCs (40 MVI-positive lesions and 52 MVI-negative lesions) were analyzed. Preoperative MR imaging examinations including diffusion kurtosis imaging (b values: 0, 200, 500, 1000, 1500, and 2000 sec/mm2) were performed and kurtosis, diffusivity, and apparent diffusion coefficient maps were calculated. Morphologic features of conventional MR images were also evaluated. Univariate and multivariate logistic regression analyses were used to evaluate the relative value of these parameters as potential predictors of MVI. Results Features significantly related to MVI of HCC at univariate analysis were increased mean kurtosis value (P < .001), decreased mean diffusivity value (P = .033) and apparent diffusion coefficient value (P = .011), and presence of infiltrative border with irregular shape (P = .005) and irregular circumferential enhancement (P = .026). At multivariate analysis, mean kurtosis value (odds ratio, 6.25; P = .001), as well as irregular circumferential enhancement (odds ratio, 6.92; P = .046), were independent risk factors for MVI of HCC. The mean kurtosis value for MVI of HCC showed an area under the receiver operating characteristic curve of 0.784 (optimal cutoff value was 0.917). Conclusion Higher mean kurtosis values in combination with irregular circumferential enhancement are potential predictive biomarkers for MVI of HCC. © RSNA, 2017.

Radiomics score: a potential prognostic imaging feature for postoperative survival of solitary HCC patients.

BACKGROUND: Radiomics is an emerging field in oncological research. In this study, we aimed at developing a radiomics score (rad-score) to estimate postoperative recurrence and survival in patients with solitary hepatocellular carcinoma (HCC). METHODS: A total of 319 solitary HCC patients (training cohort: n = 212; validation cohort: n = 107) were enrolled. Radiomics features were extracted from the artery phase of preoperatively acquired computed tomography (CT) in all patients. A rad-score was generated by using the least absolute shrinkage and selection operator (lasso) logistic model. Kaplan-Meier and Cox's hazard regression analyses were used to evaluate the prognostic significance of the rad-score. Final nomograms predicting recurrence and survival of solitary HCC patients were established based on the rad-score and clinicopathological factors. C-index and calibration statistics were used to assess the performance of nomograms. RESULTS: Six potential radiomics features were selected out of 110 texture features to formulate the rad-score. Low rad-score positively correlated with aggressive tumor phenotypes, like larger tumor size and vascular invasion. Meanwhile, low rad-score was significantly associated with increased recurrence and reduced survival. In addition, multivariate analysis identified the rad-score as an independent prognostic factor (recurrence: Hazard ratio (HR): 2.472, 95% confident interval (CI): 1.339-4.564, p = 0.004;survival: HR: 1.558, 95%CI: 1.022-2.375, p = 0.039). Notably, the nomogram integrating rad-score had a better prognostic performance as compared with traditional staging systems. These results were further confirmed in the validation cohort. CONCLUSIONS: The preoperative CT image based rad-score was an independent prognostic factor for the postoperative outcome of solitary HCC patients. This score may be complementary to the current staging system and help to stratify individualized treatments for solitary HCC patients.

Advanced imaging to predict response to chemotherapy in colorectal liver metastases - a systematic review.

BACKGROUND: The assessment of colorectal liver metastases (CRLM) after treatment with chemotherapy is challenging due to morphological and/or functional change without changes in size. The aim of this review was to assess the value of FDG-PET, FDG-PET-CT, CT and MRI in predicting response to chemotherapy in CRLM. METHODS: A systematic review was undertaken based on PRISMA statement. PubMed and Embase were searched up to October 2016 for studies on the accuracy of PET, PET-CT, CT and MRI in predicting RECIST or metabolic response to chemotherapy and/or survival in patients with CRLM. Articles evaluating the assessment of response after chemotherapy were excluded. RESULTS: Sixteen studies met the inclusion criteria and were included for further analysis. Study results were available for 6 studies for FDG-PET(-CT), 6 studies for CT and 9 studies for MRI. Generally, features predicting RECIST or metabolic response often predicted shorter survival. The ADC (apparent diffusion coefficient, on MRI) seems to be the most promising predictor of response and survival. In CT-related studies, few attenuation-related parameters and texture features show promising results. In FDG-PET(-CT), findings were ambiguous. CONCLUSION: Radiological data on the prediction of response to chemotherapy for CRLM is relatively sparse and heterogeneous. Despite that, a promising parameter might be ADC. Second, there seems to be a seemingly counterintuitive correlation between parameters that predict a good response and also predict poor survival.

Whole-tumor MRI histogram analyses of hepatocellular carcinoma: Correlations with Ki-67 labeling index.

PURPOSE: To evaluate whether whole-tumor histogram-derived parameters for an apparent diffusion coefficient (ADC) map and contrast-enhanced magnetic resonance imaging (MRI) could aid in assessing Ki-67 labeling index (LI) of hepatocellular carcinoma (HCC). MATERIALS AND METHODS: In all, 57 patients with HCC who underwent pretreatment MRI with a 3T MR scanner were included retrospectively. Histogram parameters including mean, median, standard deviation, skewness, kurtosis, and percentiles (5th , 25th , 75th , 95th ) were derived from the ADC map and MR enhancement. Correlations between histogram parameters and Ki-67 LI were evaluated and differences between low Ki-67 (≤10%) and high Ki-67 (>10%) groups were assessed. RESULTS: Mean, median, 5th , 25th , 75th percentiles of ADC, and mean, median, 25th , 75th , 95th percentiles of enhancement of arterial phase (AP) demonstrated significant inverse correlations with Ki-67 LI (rho up to -0.48 for ADC, -0.43 for AP) and showed significant differences between low and high Ki-67 groups (P < 0.001-0.04). Areas under the receiver operator characteristics (ROC) curve for identification of high Ki-67 were 0.78, 0.77, 0.79, 0.82, and 0.76 for mean, median, 5th , 25th , 75th percentiles of ADC, respectively, and 0.74, 0.81, 0.76, 0.82, 0.69 for mean, median, 25th , 75th , 95th percentiles of AP, respectively. CONCLUSION: Histogram-derived parameters of ADC and AP were potentially helpful for predicting Ki-67 LI of HCC. LEVEL OF EVIDENCE: 3 Technical Efficacy: Stage 3 J. MAGN. RESON. IMAGING 2017;46:383-392.

Improving Detection of Iron Deposition in Cirrhotic Liver Using Susceptibility-Weighted Imaging With Emphasis on Histopathological Correlation.

OBJECTIVE: The aim of this study was to investigate the value of susceptibility-weighted imaging (SWI) for detection and quantification of iron deposition in cirrhotic liver. METHODS: Fifty-five cirrhotic patients underwent hepatic magnetic resonance imaging examination including SWI and multiecho T2*-weighted imaging (T2*WI). Detection of iron deposition and number of siderotic nodules were compared between SWI and T2*WI. Correlation among SWI phase value, T2* value, and hepatic iron concentration were determined. RESULTS: Susceptibility-weighted imaging significantly improved detection of iron deposition compared with T2*WI (90.7% vs 66.7%, P = 0.002), attributing to grade 1 (73.3% vs 26.7%, P = 0.027) and grade 2 (93.8% vs 56.3%, P = 0.037). Iron deposition of grade 3 and 4 could be detected by both SWI and T2*WI. The number of siderotic nodules visualized on SWI was significantly larger than that on T2*WI (107.5 ± 7.4 vs 62.7 ± 4.6, P = 0.002). There were significantly negative correlation between phase value and iron score (r = -0.803), and positive correlation between phase value and T2* value (r = 0.771). CONCLUSIONS: Susceptibility-weighted imaging can improve detection of minimal and mild iron deposition in cirrhotic liver.

Potential of MR histogram analyses for prediction of response to chemotherapy in patients with colorectal hepatic metastases.

PURPOSE: To determine if magnetic resonance imaging (MRI) histogram analyses can help predict response to chemotherapy in patients with colorectal hepatic metastases by using response evaluation criteria in solid tumours (RECIST1.1) as the reference standard. MATERIALS AND METHODS: Standard MRI including diffusion-weighted imaging (b=0, 500 s/mm(2)) was performed before chemotherapy in 53 patients with colorectal hepatic metastases. Histograms were performed for apparent diffusion coefficient (ADC) maps, arterial, and portal venous phase images; thereafter, mean, percentiles (1st, 10th, 50th, 90th, 99th), skewness, kurtosis, and variance were generated. Quantitative histogram parameters were compared between responders (partial and complete response, n=15) and non-responders (progressive and stable disease, n=38). Receiver operator characteristics (ROC) analyses were further analyzed for the significant parameters. RESULTS: The mean, 1st percentile, 10th percentile, 50th percentile, 90th percentile, 99th percentile of the ADC maps were significantly lower in responding group than that in non-responding group (p=0.000-0.002) with area under the ROC curve (AUCs) of 0.76-0.82. The histogram parameters of arterial and portal venous phase showed no significant difference (p>0.05) between the two groups. CONCLUSION: Histogram-derived parameters for ADC maps seem to be a promising tool for predicting response to chemotherapy in patients with colorectal hepatic metastases. KEY POINTS: • ADC histogram analyses can potentially predict chemotherapy response in colorectal liver metastases. • Lower histogram-derived parameters (mean, percentiles) for ADC tend to have good response. • MR enhancement histogram analyses are not reliable to predict response.