To establish and validate a non-invasive deep learning (DL) model based on contrast-enhanced ultrasound (CEUS) to predict vessels encapsulating tumor clusters (VETC) patterns in hepatocellular carcinoma (HCC).

This retrospective study included consecutive HCC patients with preoperative CEUS images and available tissue specimens. Patients were randomly allocated into the training and test cohorts. CEUS images were analyzed using the ResNet-18 convolutional neural network for the development and validation of the VETC predictive model. The predictive value for postoperative early recurrence (ER) of the proposed model was further evaluated.

A total of 242 patients were enrolled finally, including 195 in the training cohort (54.6 ± 11.2 years, 178 males) and 47 in the test cohort (55.1 ± 10.6 years, 40 males). The DL model (DL signature) achieved favorable performance in both the training cohort (area under the receiver operating characteristics curve [AUC]: 0.92, 95% confidence interval [CI]: 0.88-0.96) and test cohort (AUC: 0.90, 95% CI: 0.82-0.99). The stratified analysis demonstrated good discrimination of DL signature regardless of tumor size. Moreover, the DL signature was found independently correlated with postoperative ER (hazard ratio [HR]: 1.99, 95% CI: 1.29-3.06, p = 0.002). C-indexes of 0.70 and 0.73 were achieved when the DL signature was used to predict ER independently and combined with clinical features.

The proposed DL signature provides a non-invasive and practical method for VETC-HCC prediction, and contributes to the identification of patients with high risk of postoperative ER.

This DL model based on contrast-enhanced US displayed an important role in non-invasive diagnosis and prognostication for patients with VETC-HCC, which was helpful in individualized management.

Preoperative biopsy to determine VETC status in HCC patients is limited. The contrast-enhanced DL model provides a non-invasive tool for the prediction of VETC-HCC. The proposed deep-learning signature assisted in identifying patients with a high risk of postoperative ER.

This study aims to investigate the clinicopathological and ultrasonography characteristics of combined hepatocellular-cholangiocarcinoma (cHCC-CCA) and its correlation with microvascular invasion (MVI), as well as the predictive value of contrast-enhanced ultrasound (CEUS) imaging.

A retrospective analysis was conducted on 57 patients diagnosed with cHCC-CCA between November 2017 and May 2023 at Guizhou Provincial People's Hospital. Among them, 27 patients were MVI-positive and 30 patients were MVI-negative, all of whom underwent preoperative CEUS within 2 weeks. Clinical data, ultrasonographic findings, and CEUS features were compared between the two groups to analyze the influencing factors and predictive value of MVI in cHCC-CCA patients.

Compared to the MVI-negative group, the MVI-positive group showed a higher proportion of tumors with a maximum diameter greater than 5 cm, elevated alpha-fetoprotein (AFP) levels, low echo halo around the tumor, non-smooth tumor contour, peripheral irregular rim-like enhancement and early washout (≤60s) with nodular patterns on CEUS (P<0.05). Multivariate logistic regression analysis revealed that low echo halo, peripheral irregular rim-like enhancement, and early washout were independent risk factors for MVI in cHCC-CCA patients. The receiver operating characteristic (ROC) curve analysis demonstrated an area under the curve (AUC) of 0.8056 for these factors.

Ultrasonographic and CEUS features have a certain correlation with MVI in cHCC-CCA patients. Low echo halo, peripheral irregular rim-like enhancement, and early washout are independent risk factors for MVI in patients with cHCC-CCA. These features have a predictive value in determining the presence of MVI in patients with cHCC-CCA.

To conduct a head-to-head comparison between deep learning (DL) and radiomics models across institutions for predicting microvascular invasion (MVI) in hepatocellular carcinoma (HCC) and to investigate the model robustness and generalizability through rigorous internal and external validation.

This retrospective study included 2304 preoperative images of 576 HCC lesions from two centers, with MVI status determined by postoperative histopathology. We developed DL and radiomics models for predicting the presence of MVI using B-mode ultrasound, contrast-enhanced ultrasound (CEUS) at the arterial, portal, and delayed phases, and a combined modality (B + CEUS). For radiomics, we constructed models with enlarged vs. original regions of interest (ROIs). A cross-validation approach was performed by training models on one center's dataset and validating the other, and vice versa. This allowed assessment of the validity of different ultrasound modalities and the cross-center robustness of the models. The optimal model combined with alpha-fetoprotein (AFP) was also validated. The head-to-head comparison was based on the area under the receiver operating characteristic curve (AUC).

Thirteen DL models and 25 radiomics models using different ultrasound modalities were constructed and compared. B + CEUS was the optimal modality for both DL and radiomics models. The DL model achieved AUCs of 0.802-0.818 internally and 0.667-0.688 externally across the two centers, whereas radiomics achieved AUCs of 0.749-0.869 internally and 0.646-0.697 externally. The radiomics models showed overall improvement with enlarged ROIs (P < 0.05 for both CEUS and B + CEUS modalities). The DL models showed good cross-institutional robustness (P > 0.05 for all modalities, 1.6-2.1% differences in AUC for the optimal modality), whereas the radiomics models had relatively limited robustness across the two centers (12% drop-off in AUC for the optimal modality). Adding AFP improved the DL models (P < 0.05 externally) and well maintained the robustness, but did not benefit the radiomics model (P > 0.05).

Cross-institutional validation indicated that DL demonstrated better robustness than radiomics for preoperative MVI prediction in patients with HCC, representing a promising solution to non-standardized ultrasound examination procedures.

The research field of artificial intelligence (AI) in medicine and especially in gastroenterology is rapidly progressing with the first AI tools entering routine clinical practice, for example, in colorectal cancer screening. Contrast-enhanced ultrasound (CEUS) is a highly reliable, low-risk, and low-cost diagnostic modality for the examination of the liver. However, doctors need many years of training and experience to master this technique and, despite all efforts to standardize CEUS, it is often believed to contain significant interrater variability. As has been shown for endoscopy, AI holds promise to support examiners at all training levels in their decision-making and efficiency.

In this systematic review, we analyzed and compared original research studies applying AI methods to CEUS examinations of the liver published between January 2010 and February 2024. We performed a structured literature search on PubMed, Web of Science, and IEEE. Two independent reviewers screened the articles and subsequently extracted relevant methodological features, e.g., cohort size, validation process, machine learning algorithm used, and indicative performance measures from the included articles.

We included 41 studies with most applying AI methods for classification tasks related to focal liver lesions. These included distinguishing benign versus malignant or classifying the entity itself, while a few studies tried to classify tumor grading, microvascular invasion status, or response to transcatheter arterial chemoembolization directly from CEUS. Some articles tried to segment or detect focal liver lesions, while others aimed to predict survival and recurrence after ablation. The majority (25/41) of studies used hand-picked and/or annotated images as data input to their models. We observed mostly good to high reported model performances with accuracies ranging between 58.6% and 98.9%, while noticing a general lack of external validation.

Even though multiple proof-of-concept studies for the application of AI methods to CEUS examinations of the liver exist and report high performance, more prospective, externally validated, and multicenter research is needed to bring such algorithms from desk to bedside.

This study aimed to determine the feasibility of using the deep learning (DL) method to determine the degree (whether myometrial invasion [MI] >50%) of MI in patients with endometrial cancer (EC) based on ultrasound (US) images.

From September 2017 to April 2023, 1289 US images of 604 patients with EC who underwent surgical resection at center 1, center 2 or center 3 were obtained and divided into a training set and an internal validation set. Ninety-five patients from center 4 and center 5 were randomly selected as the external testing set according to the same criteria as those for the primary cohort. This study evaluated three DL models trained on the training set and tested them on the validation and testing sets. The models' performance was analyzed based on accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC), and the performance of the models was subsequently compared with that of 15 radiologists.

In the final clinical diagnosis of MI in patients with EC, EfficientNet-B6 showed the best performance in the testing set in terms of area under the curve (AUC) [0.814, 95% CI (0.746-0.882]; accuracy [0.802, 95% CI (0.733-0.855]; sensitivity [0.623]; specificity [0.879]; positive likelihood ratio (PLR) [6.750]; and negative likelihood ratio (NLR) [0.389]. The diagnostic efficacy of EfficientNet-B6 was significantly better than that of the 15 radiologists, with an average diagnostic accuracy of 0.681, average AUC of 0.678, AUC of the best performance of 0.739, accuracy of 0.716, sensitivity of 0.806, specificity 0.672, PLR2.457, and NLR 0.289.

Based on the preoperative US images of patients with EC, the DL model can accurately determine the degree of endometrial MI; the performance of this model is significantly better than that of radiologists, and it can effectively assist in clinical treatment decisions.

To exploit the improved prediction performance based on dynamic contrast-enhanced (DCE) MRI by using dynamic radiomics for microvascular invasion (MVI) in hepatocellular carcinoma (HCC).

We retrospectively included 175 and 75 HCC patients who underwent preoperative DCE-MRI from September 2019 to August 2022 in institution 1 (development cohort) and institution 2 (validation cohort), respectively. Static radiomics features were extracted from the mask, arterial, portal venous, and equilibrium phase images and used to construct dynamic features. The static, dynamic, and dynamic-static radiomics (SR, DR, and DSR) signatures were separately constructed based on the feature selection method of LASSO and classification algorithm of logistic regression. The receiver operating characteristic (ROC) curves and the area under the curve (AUC) were plotted to evaluate and compare the predictive performance of each signature.

In the three radiomics signatures, the DSR signature performed the best. The AUCs of the SR, DR, and DSR signatures in the training set were 0.750, 0.751 and 0.805, respectively, while in the external validation set, the corresponding AUCs were 0.706, 0756 and 0.777. The DSR signature showed significant improvement over the SR signature in predicting MVI status (training cohort: P = 0.019; validation cohort: P = 0.044). After external validation, the AUC value of the SR signature decreased from 0.750 to 0.706, while the AUC value of the DR signature did not show a decline (AUCs: 0.756 vs. 0.751).

The dynamic radiomics had an improved effect on the MVI prediction in HCC, compared with the static DCE MRI-based radiomics models.

To describe the frequency of open science practices in a contemporary sample of studies developing prognostic models using machine learning methods in the field of oncology.

We conducted a systematic review, searching the MEDLINE database between December 1, 2022, and December 31, 2022, for studies developing a multivariable prognostic model using machine learning methods (as defined by the authors) in oncology. Two authors independently screened records and extracted open science practices.

We identified 46 publications describing the development of a multivariable prognostic model. The adoption of open science principles was poor. Only one study reported availability of a study protocol, and only one study was registered. Funding statements and conflicts of interest statements were common. Thirty-five studies (76%) provided data sharing statements, with 21 (46%) indicating data were available on request to the authors and seven declaring data sharing was not applicable. Two studies (4%) shared data. Only 12 studies (26%) provided code sharing statements, including 2 (4%) that indicated the code was available on request to the authors. Only 11 studies (24%) provided sufficient information to allow their model to be used in practice. The use of reporting guidelines was rare: eight studies (18%) mentioning using a reporting guideline, with 4 (10%) using the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis Or Diagnosis statement, 1 (2%) using Minimum Information About Clinical Artificial Intelligence Modeling and Consolidated Standards Of Reporting Trials-Artificial Intelligence, 1 (2%) using Strengthening The Reporting Of Observational Studies In Epidemiology, 1 (2%) using Standards for Reporting Diagnostic Accuracy Studies, and 1 (2%) using Transparent Reporting of Evaluations with Nonrandomized Designs.

The adoption of open science principles in oncology studies developing prognostic models using machine learning methods is poor. Guidance and an increased awareness of benefits and best practices of open science are needed for prediction research in oncology.

This study aimed to develop and validate a combined nomogram model based on superb microvascular imaging (SMI)-based deep learning (DL), radiomics characteristics, and clinical factors for noninvasive differentiation between immunoglobulin A nephropathy (IgAN) and non-IgAN.We prospectively enrolled patients with chronic kidney disease who underwent renal biopsy from May 2022 to December 2022 and performed an ultrasound and SMI the day before renal biopsy. The selected patients were randomly divided into training and testing cohorts in a 7:3 ratio. We extracted DL and radiometric features from the two-dimensional ultrasound and SMI images. A combined nomograph model was developed by combining the predictive probability of DL with clinical factors using multivariate logistic regression analysis. The proposed model's utility was evaluated using receiver operating characteristics, calibration, and decision curve analysis. In this study, 120 patients with primary glomerular disease were included, including 84 in the training and 36 in the test cohorts. In the testing cohort, the ROC of the radiomics model was 0.816 (95% CI:0.663-0.968), and the ROC of the DL model was 0.844 (95% CI:0.717-0.971). The nomogram model combined with independent clinical risk factors (IgA and hematuria) showed strong discrimination, with an ROC of 0.884 (95% CI:0.773-0.996) in the testing cohort. Decision curve analysis verified the clinical practicability of the combined nomogram. The combined nomogram model based on SMI can accurately and noninvasively distinguish IgAN from non-IgAN and help physicians make clearer patient treatment plans.

Microvascular invasion (MVI) is an important predictor of poor prognosis in patients with hepatocellular carcinoma (HCC). Accurate preoperative prediction of MVI in HCC would provide useful information to guide the choice of therapeutic strategy. Shear wave elastography (SWE) plays an important role in hepatic imaging, but its value in the preoperative prediction of MVI in HCC has not yet been proven.

To explore the value of conventional ultrasound features and SWE in the preoperative prediction of MVI in HCC.

Patients with a postoperative pathological diagnosis of HCC and a definite diagnosis of MVI were enrolled in this study. Conventional ultrasound features and SWE features such as maximal elasticity (Emax) of HCCs and Emax of the periphery of HCCs were acquired before surgery. These features were compared between MVI-positive HCCs and MVI-negative HCCs and between mild MVI HCCs and severe MVI HCCs.

This study included 86 MVI-negative HCCs and 102 MVI-positive HCCs, including 54 with mild MVI and 48 with severe MVI. Maximal tumor diameters, surrounding liver tissue, color Doppler flow, Emax of HCCs, and Emax of the periphery of HCCs were significantly different between MVI-positive HCCs and MVI-negative HCCs. In addition, Emax of the periphery of HCCs was significantly different between mild MVI HCCs and severe MVI HCCs. Higher Emax of the periphery of HCCs and larger maximal diameters were independent risk factors for MVI, with odds ratios of 2.820 and 1.021, respectively.

HCC size and stiffness of the periphery of HCC are useful ultrasound criteria for predicting positive MVI. Preoperative ultrasound and SWE can provide useful information for the prediction of MVI in HCCs.

The aim of this study was to establish and validate a contrast-enhanced ultrasound (CEUS) nomogram for pre-operative microvascular invasion (MVI) prediction in hepatocellular carcinoma (HCC), and compare it with the nomogram based on gadopentetate dimeglumine-enhanced magnetic resonance imaging (Gd-MRI).

A total of 251 patients with a single HCC were enrolled in this prospective study, including 176 patients in the training cohort and 75 patients in the validation cohort. Contrast-enhanced ultrasound (CEUS) with Sonazoid and Gd-MRI was performed pre-operatively. Post-operative histopathology was the gold standard for MVI. Univariate and multivariate logistic regression was performed to determine independent risk factors for MVI. Nomograms based on CEUS and Gd-MRI were established, and their discrimination, calibration and decision curve analysis were evaluated and compared.

Multivariate logistic regression revealed that arterial circular enhancement, non-enhancing area and thick ring-like enhancement in the post-vascular phase were independent risk factors for MVI. The areas under the receiver operating characteristic curve of the nomogram were 0.841 (0.779-0.892) and 0.914 (0.827-0.966) in the training and validation cohorts, with no significant difference compared with the Gd-MRI nomogram (p = 0.294, 0.321). The C-indexes were 0.821 and 0.870 in the training and validation cohorts. Decision curve analysis revealed that the CEUS nomogram had better clinical applicability than the Gd-MRI nomogram when the threshold probability was between 0.35 and 0.95.

The CEUS-based nomogram was available for predicting MVI in HCC, and its predictive performance was not inferior to that of Gd-MRI.