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.

Time-of-Flight Intracranial MRA at 3 T versus 5 T versus 7 T: Visualization of Distal Small Cerebral Arteries.

Background Three-dimensional (3D) time-of-flight (TOF) MR angiography (MRA) at 7 T has been reported to have high image quality for visualizing small perforating vessels. However, B1 inhomogeneity and more physiologic considerations limit its applications. Angiography at 5 T may provide another choice for intracranial vascular imaging. Purpose To evaluate the image quality and cerebrovascular visualization of 5-T 3D TOF MRA for visualizing intracranial small branch arteries. Materials and Methods Participants (healthy volunteers or participants with a history of ischemic stroke undergoing intracranial CT angiography or MRA for identifying steno-occlusive disease) were prospectively included from September 2021 to November 2021. Each participant underwent 3-T, 5-T, and 7-T 3D TOF MRA with use of customized MR protocols within 48 hours. Radiologist scoring from 0 (invisible) to 3 (excellent) and quantitative assessment were obtained to evaluate the image quality. The Friedman test was used for comparison of characteristics derived from 3 T, 5 T, and 7 T. Results A total of 12 participants (mean age ± SD, 38 years ± 9; nine men) were included. Visualizations of the distal arteries and small vessels at 5-T TOF MRA were significantly higher than those at 3 T (median score: 3.0 vs 2.0, all P < .001 for distal segments and lenticulostriate artery; median score: 2.0 vs 0, P < .001 for pontine artery). The total length of small vessel branches detected at 5 T was larger than that at 3 T (5.1 m ± 0.7 vs 1.9 m ± 0.4; P < .001). However, there was no evidence of a significant difference compared with 7 T in either the depiction of distal segments and small vessel branches (average median score, 2.5; all P > .05) or the quantitative measurements (total length, 5.6 m ± 0.5; P = .41). Conclusion Three-dimensional time-of-flight MR angiography at 5 T presented the capability to provide superior visualization of distal large arteries and small vessel branches (in terms of subjective and quantitative assessment) to 3 T and had image quality similar to 7 T. © RSNA, 2022 Online supplemental material is available for this article. An earlier incorrect version appeared online. This article was corrected on September 14, 2022.

Characterization of Microvascular Invasion in Hepatocellular Carcinoma Using Computational Modeling of Interstitial Fluid Pressure and Velocity.

BACKGROUND: Most solid tumors show increased interstitial fluid pressure (IFP), and this increased IFP is an obstacle to treatment. A noninvasive model for measuring IFP in hepatocellular carcinoma (HCC) is an unresolved issue. PURPOSE: To develop a noninvasive model to measure IFP and interstitial fluid velocity (IFV) in HCC and to characterize the microvascular invasion (MVI) status by using this model. STUDY TYPE: Retrospective. POPULATION: A total of 97 HCC patients (mean age 57.6 ± 10.9 years, 77.3% males), 53 of them with MVI and 44 of them without MVI. FIELD STRENGTH/SEQUENCE: A 3-T, three-dimensional spoiled gradient-recalled echo. ASSESSMENT: MVI was defined as microscopic vascular invasion of small vessels within the peritumoral liver tissue. The volumes of interest (VOIs) were manually delineated and enclosed the tumor lesion and healthy liver parenchyma, respectively. The extended Tofts model (ETM) was used to estimate permeability parameters from all the VOIs. Subsequently, the continuity partial differential equation (PDE) was implemented and IFP and IFV were acquired. STATISTICAL TESTS: Wilcoxon signed-ranks tests, histogram analysis, Mann-Whitney U test, Fisher's exact test, least absolute shrinkage and selection operator (LASSO) logistic regression, receiver operating characteristic (ROC) curve analysis with the area under the curve (AUC), Youden index, DeLong test, and Benjamini-Hochberg correction. A P value <0.05 was considered statistically significant. RESULTS: The HCC lesions exhibited elevated IFP and reduced IFV. There were no significant differences in any measured demographic and clinical features between the MVI-positive and MVI-negative groups, except for tumor size. Nine IFP histogram analysis-derived parameters and seven IFV histogram analysis-derived parameters could be used to characterize the MVI status. LASSO regression selected five features: IFP maximum, IFP 10th percentile, IFP 90th percentile, IFV SD, and IFV 10th percentile. The combination of these features showed the highest AUC (0.781) and specificity (77.3%). DATA CONCLUSION: A noninvasive IFP and IFV measurement model for HCC was developed. Specific IFP- and IFV-derived parameters exhibited significant association with the MVI status. EVIDENCE LEVEL: 3. TECHNICAL EFFICACY: Stage 2.

Risk stratification for 1- to 2-cm gastric gastrointestinal stromal tumors: visual assessment of CT and EUS high-risk features versus CT radiomics analysis.

OBJECTIVES: To investigate the ability of CT and endoscopic sonography (EUS) in predicting the malignant risk of 1-2-cm gastric gastrointestinal stromal tumors (gGISTs) and to clarify whether radiomics could be applied for risk stratification. METHODS: A total of 151 pathologically confirmed 1-2-cm gGISTs from seven institutions were identified by contrast-enhanced CT scans between January 2010 and March 2021. A detailed description of EUS morphological features was available for 73 gGISTs. The association between EUS or CT high-risk features and pathological malignant potential was evaluated. gGISTs were randomly divided into three groups to build the radiomics model, including 74 in the training cohort, 37 in validation cohort, and 40 in testing cohort. The ROIs covering the whole tumor volume were delineated on the CT images of the portal venous phase. The Pearson test and least absolute shrinkage and selection operator (LASSO) algorithm were used for feature selection, and the ROC curves were used to evaluate the model performance. RESULTS: The presence of EUS- and CT-based morphological high-risk features, including calcification, necrosis, intratumoral heterogeneity, irregular border, or surface ulceration, did not differ between very-low and intermediate risk 1-2-cm gGISTs (p > 0.05). The radiomics model consisting of five radiomics features showed favorable performance in discrimination of malignant 1-2-cm gGISTs, with the AUC of the training, validation, and testing cohort as 0.866, 0.812, and 0.766, respectively. CONCLUSIONS: Instead of CT- and EUS-based morphological high-risk features, the CT radiomics model could potentially be applied for preoperative risk stratification of 1-2-cm gGISTs. KEY POINTS: • The presence of EUS- and CT-based morphological high-risk factors, including calcification, necrosis, intratumoral heterogeneity, irregular border, or surface ulceration, did not correlate with the pathological malignant potential of 1-2-cm gGISTs. • The CT radiomics model could potentially be applied for preoperative risk stratification of 1-2-cm gGISTs.

Preoperative Local Staging of Colon Cancer by CT: Radiological Staging Criteria Based on Membrane Anatomy and Visceral Adipose Tissue.

BACKGROUND: Accuracy of preoperative T staging for colon cancer remains disappointing. OBJECTIVE: This study aimed to propose specially designed radiological staging criteria based on membrane anatomy and visceral adipose tissue and compare the staging performance with the routinely used method. DESIGN: This is a prospective observational study. SETTING: This study was conducted at a high-volume colorectal center. PARTICIPANTS: Consecutive patients with colonoscopy-proven colon carcinoma referred for clinical staging and elective resection were enrolled. INTERVENTION: The preoperative CT data were separately reviewed by 2 teams of radiologists for assigning T-stage categories (T1-2, T3, or T4) using the routine staging method or the newly proposed radiological criteria. MEASURES: Diagnostic performance for T staging was compared between the 2 criteria. RESULTS: Between October 2019 and August 2020, 190 patients were included. Compared with pathological results, T stage was correctly determined in 113 of 190 patients (59.5%) with the conventional CT criteria. With the newly developed criteria, 160 patients (84.2%) were found to be correctly staged. Accuracies between the 2 criteria significantly differed ( p < 0.001). For T1-2 staging, there were no significant differences between the sensitivities of conventional and new criteria (57.1% vs 61.9%; p = 0.990) or between their specificities (95.3% vs 98.2%; p = 0.131). However, for T3 and T4 staging, the newly developed CT criteria exhibited significantly higher sensitivity (T3: 85.2% vs 57.4%; p < 0.001; T4: 90.7% vs 64.8%; p < 0.001) and specificity (T3: 82.7% vs 64%; p = 0.006; T4: 89.7% vs 69.1%; p < 0.001) than the conventional criteria. Moreover, the new criteria (area under the curve = 0.902) performed significantly better than the conventional criteria (area under the curve = 0.670; p < 0.001), for identifying the T4-stage tumor. LIMITATIONS: The limitations are that it is a single-center study and there was no external validation. CONCLUSIONS: The specially designed radiological criteria can offer more accurate T staging than the routine method in colon cancer. See Video Abstract at http://links.lww.com/DCR/B992 . PREDICCIN DE LA MORTALIDAD A DAS POSTERIORES A LA PRIMERA CIRUGA EN PACIENTES CON CNCER DE COLON OBSTRUCTIVO DEL LADO IZQUIERDO: ANTECEDENTES:Se cree que la resección aguda para el carcinoma de colon obstructivo del lado izquierdo está asociada con un mayor riesgo de mortalidad que un enfoque puente a la cirugía que utiliza un estoma de descompresión o un stent metálico autoexpandible, pero faltan modelos de predicción.OBJETIVO:Determinar la influencia de la estrategia de tratamiento sobre la mortalidad dentro de los 90 días desde la primera intervención utilizando un modelo de predicción en pacientes que presentan carcinoma de colon obstructivo del lado izquierdo.DISEÑO:Un estudio de cohorte multicéntrico nacional, utilizando datos de una auditoría nacional prospectiva.ENTORNO CLINICO:El estudio se realizó en 75 hospitales holandeses.PACIENTES:Se incluyeron los pacientes que se sometieron a una resección con intención curativa de un carcinoma de colon obstructivo del lado izquierdo entre 2009 y 2016.INTERVENCIONES:La primera intervención fue resección aguda, puente a cirugía con stent metálico autoexpandible o puente a cirugía con estoma descompresor.PRINCIPALES MEDIDAS DE VALORACIÓN:La principal medida de resultado fue la mortalidad a los 90 días después de la primera intervención. Los factores de riesgo se identificaron mediante análisis logístico multivariable. Posteriormente se desarrolló un modelo de riesgo.RESULTADOS:En total se incluyeron 2395 pacientes, siendo la primera intervención resección aguda en 1848 (77%) pacientes, estoma como puente a la cirugía en 332 (14%) pacientes y stent como puente a la cirugía en 215 (9%) pacientes. En general, 152 pacientes (6,3%) fallecieron dentro de los 90 días posteriores a la primera intervención. Un estoma de descompresión se asoció de forma independiente con un menor riesgo de mortalidad a los 90 días (HR: 0,27, IC: 0,094-0,62). Otros predictores independientes de mortalidad fueron la edad, la clasificación ASA, la ubicación del tumor y los niveles índice de creatinina sérica y proteína C reactiva. El modelo de riesgo construido tuvo un área bajo la curva de 0,84 (IC: 0,81-0,87).LIMITACIONES:Solo se incluyeron pacientes que se sometieron a resección quirúrgica.CONCLUSIONES:La estrategia de tratamiento tuvo un impacto significativo en la mortalidad a los 90 días. Un estoma descompresor reduce considerablemente el riesgo de mortalidad, especialmente en pacientes mayores y frágiles. Se desarrolló un modelo de riesgo, que necesita una mayor validación externa. Consulte Video Resumen en http://links.lww.com/DCR/B992 . (Traducción-Dr. Ingrid Melo ).

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.

Preliminary Exploration of the Application of Vesical Imaging-Reporting and Data System (VI-RADS) in Post-treatment Patients With Bladder Cancer: A Prospective Single-Center Study.

BACKGROUND: Vesical Imaging-Reporting and Data System (VI-RADS) has been shown to be effective in diagnosing muscle invasion of bladder cancer (BC) in primary patients. PURPOSE: To evaluate the diagnostic efficacy of VI-RADS in a BC target population which included post-treatment patients, and to determine the repeatability. STUDY TYPE: Prospective. POPULATION: Seventy-three patients (42 with primary BC, 31 with post-treatment BC). FIELD STRENGTH/SEQUENCE: 3.0 T MRI with propeller fast spin-echo T2 WI, echo planer imaging diffusion-weighted imaging (DWI), and dynamic contrast-enhanced imaging (DCEI). ASSESSMENT: VI-RADS scores were independently assessed by five radiologists with different levels of experience. The diagnostic efficiency in each group (primary and post-treatment) and of each radiologist was assessed. STATISTICAL TESTS: Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), and area under the curve (AUC) in receiver operating characteristic curve analysis were calculated to evaluate VI-RADS diagnostic performance. Interobserver agreement was assessed using weighted Kappa statistics. A P value <0.05 was considered statistically significant. RESULTS: At the corresponding cut-off, AUC values of three groups range from 0.936 to 0.947 and AUC values of five observers range from 0.901 to 0.963. There was no significant difference between the AUCs in the primary and post-treatment groups (P = 0.870). The cut-off of the whole group and the post-treatment group was ≥4, and the cut-off of the primary group was ≥3. The Kappa values of interobserver agreements range from 0.709 to 0.923. CONCLUSIONS: After expanding the target population to include post-treatment patients, VI-RADS still has good diagnostic efficacy and repeatability. VI-RADS could potentially be a preoperative staging tool for post-treatment patients. LEVEL OF EVIDENCE: 1 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.

T1 Mapping on Gd-EOB-DTPA-Enhanced MRI for the Prediction of Oxaliplatin-Induced Liver Injury in a Mouse Model.

BACKGROUND: Oxaliplatin-induced liver injury (OILI) not only impairs hepatic regeneration but also increases postoperative morbidity and mortality. Therefore, noninvasive, accurate, and early diagnosis of OILI is mandatory. PURPOSE: To evaluate the potential of T1 mapping on gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced MRI for assessing OILI in a mouse model. STUDY TYPE: Case control, animal model. ANIMAL MODEL: Thirty oxaliplatin-treated mice and 10 control mice were included. FIELD STRENGTH: Volumetric interpolated breath-hold examination sequence: 3T scanner with a phased-array animal 8-channel coil. T1 mapping before and at hepatobiliary phase (HBP) after injection of Gd-EOB-DTPA were undertaken. ASSESSMENT: T1 relaxation times of the liver parenchyma were measured and the reduction rate (ΔT1 %) was calculated. Histological findings were used as a standard reference. STATISTICAL TESTS: The Kruskal-Wallis test with pairwise comparisons using the Mann-Whitney U-test were applied to compare the parameters across groups. Spearman's rank correlation test and receiver operating characteristics (ROC) analyses were performed. Areas under the curves (AUCs) were compared using the DeLong method. RESULTS: Histologically, mice were classified as normal (n = 10), hepatocellular degeneration without fibrosis (n = 16), and hepatocellular degeneration with fibrosis (n = 14). HBP T1 relaxation time increased with the severity of OILI (rho = 0.60, P < 0.05), and ΔT1 % decreased with the severity of OILI (rho = -0.78, P < 0.05). AUC was 0.92 for ΔT1 % in differentiating hepatocellular degeneration without fibrosis from normal liver, but HBP T1 relaxation time could not distinguish them (P = 0.09). AUCs were 0.96 and 0.95 for HBP T1 relaxation time, and 0.90 and 0.84 for ΔT1 % in discriminating OILI with fibrosis from normal liver and OILI without fibrosis. DATA CONCLUSION: HBP T1 relaxation time and ΔT1 % of Gd-EOB-DTPA enhanced MRI was useful for assessing OILI. ΔT1 % may be more sensitive than HBP T1 relaxation time in detecting early stage of liver injury. LEVEL OF EVIDENCE: 2. TECHNICAL EFFICACY STAGE: 5.

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.

Performance of liver stiffness measurements obtained with FibroScan is affected by glucose metabolism in patients with nonalcoholic fatty liver disease.

BACKGROUND: The performance of liver stiffness measurements (LSMs) obtained using FibroScan can be affected by several factors, and cut-off values are different for fibrosis caused by various aetiologies. The study aims to evaluate the diagnostic accuracy of LSM in nonalcoholic fatty liver disease (NAFLD) patients with abnormal glucose metabolism and investigate whether the LSM value would be affected by metabolic indicators. METHODS: The study involved 91 NAFLD patients with abnormal glucose metabolism who underwent liver biopsy. The diagnostic accuracy of LSM value was evaluated by the receiver operator characteristic (ROC) curves, with the biopsy results taken as the gold standard. Multivariate linear regression and subgroup analysis were performed to determine the correlated indicators. RESULTS: The areas under the ROC curves (AUROCs) of LSM values for detecting fibrosis stage ≥1, 2, 3 and 4 were 0.793 (95% confidence interval [CI]: 0.695-0.871), 0.764 (95% CI: 0.663-0.846), 0.837 (95% CI: 0.744-0.906) and 0.902 (95% CI: 0.822-0.955), with cut-off values of 6.3, 7.6, 8.3 and 13.8 kPa, respectively. Multivariate linear regression demonstrated that haemoglobin A1c (HbA1c, ß = 0.205, P = 0.026) and alanine aminotransferase (ALT, ß = 0.192, P = 0.047) were independently associated with the LSM value after adjustment for fibrosis stage, ballooning and inflammation grade from liver biopsy. Subgroup analysis demonstrated that LSM values were slightly higher in patients with HbA1c ≥7% than in those with HbA1c < 7% and in patients with body mass index (BMI) ≥30 kg/m2 than in those with BMI < 30 kg/m2. CONCLUSIONS: FibroScan was valuable for the evaluation of liver fibrosis in NAFLD patients with abnormal glucose metabolism. FibroScan is recommended to evaluate severe fibrosis, especially to exclude advanced fibrosis. Glucose metabolism state may affect LSM values.

Whole-tumour evaluation with MRI and radiomics features to predict the efficacy of S-1 for adjuvant chemotherapy in postoperative pancreatic cancer patients: a pilot study.

BACKGROUND: Multiple guidelines for pancreatic ductal adenocarcinoma (PDAC) suggest that all stages of patients need to receive postoperative adjuvant chemotherapy. S-1 is a recently emerged oral antitumour agent recommended by the guidelines. However, which population would benefit from S-1 needs to be determined, and predictors of chemotherapy response are needed for personalized precision medicine. This pilot study aimed to initially identify whether whole-tumour evaluation with MRI and radiomics features could be used for predicting the efficacy of S-1 and to find potential predictors of the efficacy of S-1 as evidence to assist personalized precision treatment. METHODS: Forty-six patients with PDAC (31 in the primary cohort and 15 in the validation cohort) who underwent curative resection and subsequently adjuvant chemotherapy with S-1 were included. Pre-operative abdominal contrast-enhanced MRI was performed, and radiomics features of the whole PDAC were extracted from the primary cohort. After univariable analysis and radiomics features selection, a multivariable Cox regression model for survival analysis was subsequently used to select statistically significant factors associated with postoperative disease-free survival (DFS). Predictive capacities of the factors were tested on the validation cohort by using Kaplan-Meier method. RESULTS: Multivariable Cox regression analysis identified the probability of T1WI_NGTDM_Strength and tumour location as independent predictors of the efficacy of S-1 for adjuvant chemotherapy of PDAC (p = 0.005 and 0.013) in the primary cohort, with hazard ratios (HRs) of 0.289 and 0.293, respectively. Further survival analysis showed that patients in the low-T1WI_NGTDM_Strength group had shorter DFS (median = 5.1 m) than those in the high-T1WI_NGTDM_Strength group (median = 13.0 m) (p = 0.006), and patients with PDAC on the pancreatic head exhibited shorter DFS (median = 7.0 m) than patients with tumours in other locations (median = 20.0 m) (p = 0.016). In the validation cohort, the difference in DFS between patients with low-T1WI_NGTDM_Strength and high-T1WI_NGTDM_Strength and the difference between patients with PDAC on the pancreatic head and that in other locations were approved, with marginally significant (p = 0.073 and 0.050), respectively. CONCLUSIONS: Whole-tumour radiomics feature of T1WI_NGTDM_Strength and tumour location were potential predictors of the efficacy of S-1 and for the precision selection of S-1 as adjuvant chemotherapy regimen for PDAC.

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.

Efficacy of exenatide and insulin glargine on nonalcoholic fatty liver disease in patients with type 2 diabetes.

BACKGROUND: The aim of this study was to investigate the efficacy of exenatide and insulin glargine in patients with newly diagnosed type 2 diabetes mellitus (T2DM) and nonalcoholic fatty liver disease (NAFLD). METHODS: We performed a 24-week randomized controlled multicentre clinical trial. Seventy-six patients were randomly assigned 1:1 to receive exenatide or insulin glargine treatment. The endpoints included changes in liver fat content (LFC), visceral adipose tissue (VAT), and subcutaneous adipose tissue (SAT) measured by magnetic resonance spectroscopy, blood glucose, liver enzymes, lipid profile, body weight, and Fibrosis-4 index (FIB-4). RESULTS: LFC, VAT, SAT, and FIB-4 were significantly reduced after exenatide treatment (ΔLFC, -17.55 ± 12.93%; ΔVAT, -43.57 ± 68.20 cm2 ; ΔSAT, -28.44 ± 51.48 cm2 ; ΔFIB-4, -0.10 ± 0.26; all P < .05). In comparison, only LFC (ΔLFC, -10.49 ± 11.38%; P < .05), and not VAT, SAT, or FIB-4 index (all P > .05), was reduced after insulin glargine treatment. Moreover, exenatide treatment resulted in greater reductions in alanine transaminase (ALT), aspartate transaminase (AST), and gamma glutamyl transpeptidase (GGT) than insulin glargine (P < 0.05). The body weight, waist circumference, postprandial plasma glucose, and low-density lipoprotein cholesterol (LDL-C) in the exenatide group also presented greater reductions than the insulin glargine group (P < .05). The proportion of adverse events were comparable between the two groups. CONCLUSION: Both exenatide and insulin glargine reduced LFC in patients with drug-naive T2DM and NAFLD; however, exenatide showed greater reductions in body weight, visceral fat area, liver enzymes, FIB-4, postprandial plasma glucose, and LDL-C.