Recent advances in non-invasive magnetic resonance imaging assessment of hepatocellular carcinoma

Radiomics analysis of R2* maps to predict early recurrence of single hepatocellular carcinoma after hepatectomy

Objectives

This study aimed to evaluate the effectiveness of radiomics analysis with R2* maps in predicting early recurrence (ER) in single hepatocellular carcinoma (HCC) following partial hepatectomy.

Methods

We conducted a retrospective analysis involving 202 patients with surgically confirmed single HCC having undergone preoperative magnetic resonance imaging between 2018 and 2021 at two different institutions. 126 patients from Institution 1 were assigned to the training set, and 76 patients from Institution 2 were assigned to the validation set. A least absolute shrinkage and selection operator (LASSO) regularization was conducted to operate a logistic regression, then features were identified to construct a radiomic score (Rad-score). Uni- and multi-variable tests were used to assess the correlations of clinicopathological features and Rad-score with ER. We then established a combined model encompassing the optimal Rad-score and clinical-pathological risk factors. Additionally, we formulated and validated a predictive nomogram for predicting ER in HCC. The nomogram's discrimination, calibration, and clinical utility were thoroughly evaluated.

Results

Multivariable logistic regression revealed the Rad-score, microvascular invasion (MVI), and α fetoprotein (AFP) level > 400 ng/mL as significant independent predictors of ER in HCC. We constructed a nomogram based on these significant factors. The areas under the receiver operator characteristic curve of the nomogram and precision-recall curve were 0.901 and 0.753, respectively, with an F1 score of 0.831 in the training set. These values in the validation set were 0.827, 0.659, and 0.808.

Conclusion

The nomogram that integrates the radiomic score, MVI, and AFP demonstrates high predictive efficacy for estimating the risk of ER in HCC. It facilitates personalized risk classification and therapeutic decision-making for HCC patients.

Current Imaging Diagnosis of Hepatocellular Carcinoma

Simple Summary

The role of imaging in the management of hepatocellular carcinoma (HCC) has significantly evolved and expanded beyond the plain radiological confirmation of the tumor based on the typical appearance in a multiphase contrast-enhanced CT or MRI examination. The introduction of hepatobiliary contrast agents has enabled the diagnosis of hepatocarcinogenesis at earlier stages, while the application of ultrasound contrast agents has drastically upgraded the role of ultrasound in the diagnostic algorithms. Newer quantitative techniques assessing blood perfusion on CT and MRI not only allow earlier diagnosis and confident differentiation from other lesions, but they also provide biomarkers for the evaluation of treatment response. As distinct HCC subtypes are identified, their correlation with specific imaging features holds great promise for estimating tumor aggressiveness and prognosis. This review presents the current role of imaging and underlines its critical role in the successful management of patients with HCC.

Abstract

Hepatocellular carcinoma (HCC) is the fourth leading cause of cancer related death worldwide. Radiology has traditionally played a central role in HCC management, ranging from screening of high-risk patients to non-invasive diagnosis, as well as the evaluation of treatment response and post-treatment follow-up. From liver ultrasonography with or without contrast to dynamic multiple phased CT and dynamic MRI with diffusion protocols, great progress has been achieved in the last decade. Throughout the last few years, pathological, biological, genetic, and immune-chemical analyses have revealed several tumoral subtypes with diverse biological behavior, highlighting the need for the re-evaluation of established radiological methods. Considering these changes, novel methods that provide functional and quantitative parameters in addition to morphological information are increasingly incorporated into modern diagnostic protocols for HCC. In this way, differential diagnosis became even more challenging throughout the last few years. Use of liver specific contrast agents, as well as CT/MRI perfusion techniques, seem to not only allow earlier detection and more accurate characterization of HCC lesions, but also make it possible to predict response to treatment and survival. Nevertheless, several limitations and technical considerations still exist. This review will describe and discuss all these imaging modalities and their advances in the imaging of HCC lesions in cirrhotic and non-cirrhotic livers. Sensitivity and specificity rates, method limitations, and technical considerations will be discussed.

Role of gadoxetic acid-enhanced liver magnetic resonance imaging in the evaluation of hepatocellular carcinoma after locoregional treatment

Locoregional treatments, as alternatives to surgery, play a key role in the management of hepatocellular carcinoma (HCC). Liver magnetic resonance imaging (MRI) enables a multiparametric assessment, going beyond the traditional dynamic computed tomography approach. Moreover, the use of hepatobiliary agents can improve diagnostic accuracy and are becoming important in the diagnosis and follow-up of HCC. However, the main challenge is to quickly identify classical responses to loco-regional treatments in order to determine the most suitable management strategy for each patient. The aim of this review is to provide a summary of the most common and uncommon liver MRI findings in patients who underwent loco-regional treatments for HCC, with a special focus on ablative therapies (radiofrequency, microwaves and cryoablation), trans-arterial chemoembolization, trans-arterial radio-embolization and stereotactic ablative radiotherapy techniques, considering the usefulness of gadoxetate disodium (Gd-EOB-DTPA) contrast agent.

Benign focal liver lesions: The role of magnetic resonance imaging

Liver lesions are common findings in radiologists’ daily routine. They are a complex category of pathology that range from solitary benign lesions to primary liver cancer and liver metastases. Benign focal liver lesions can arise from different liver cell types: Epithelial (hepatocytes and biliary cells) and nonepithelial (mesenchymal cells). Liver magnetic resonance imaging (MRI) is a fundamental radiological method in these patients as it allows with its multiparametric approach optimal non-invasive tissue characterization. Furthermore, advanced liver MRI techniques such as diffusion-weighted imaging and hepatobiliary contrast agents have improved the detection of focal liver lesions and can be highly effective in differentiating pseudotumor from tumors, as well as benign from malignant lesions, and can also be used for differential diagnosis. Although histological examination can be useful in making a definitive diagnosis, MRI is an important modality in the diagnosis of liver lesions with a significant impact on patient care. This aim of this review is to provide a comprehensive overview of benign liver lesions on MRI.

Radiomics Analysis Based on Contrast-Enhanced MRI for Prediction of Therapeutic Response to Transarterial Chemoembolization in Hepatocellular Carcinoma

Purpose

To investigate the role of contrast-enhanced magnetic resonance imaging (CE-MRI) radiomics for pretherapeutic prediction of the response to transarterial chemoembolization (TACE) in patients with hepatocellular carcinoma (HCC).

Methods

One hundred and twenty-two HCC patients (objective response, n = 63; non-response, n = 59) who received CE-MRI examination before initial TACE were retrospectively recruited and randomly divided into a training cohort (n = 85) and a validation cohort (n = 37). All HCCs were manually segmented on arterial, venous and delayed phases of CE-MRI, and total 2367 radiomics features were extracted. Radiomics models were constructed based on each phase and their combination using logistic regression algorithm. A clinical-radiological model was built based on independent risk factors identified by univariate and multivariate logistic regression analyses. A combined model incorporating the radiomics score and selected clinical-radiological predictors was constructed, and the combined model was presented as a nomogram. Prediction models were evaluated by receiver operating characteristic curves, calibration curves, and decision curve analysis.

Results

Among all radiomics models, the three-phase radiomics model exhibited better performance in the training cohort with an area under the curve (AUC) of 0.838 (95% confidence interval (CI), 0.753 - 0.922), which was verified in the validation cohort (AUC, 0.833; 95% CI, 0.691 - 0.975). The combined model that integrated the three-phase radiomics score and clinical-radiological risk factors (total bilirubin, tumor shape, and tumor encapsulation) showed excellent calibration and predictive capability in the training and validation cohorts with AUCs of 0.878 (95% CI, 0.806 - 0.950) and 0.833 (95% CI, 0.687 - 0.979), respectively, and showed better predictive ability (P = 0.003) compared with the clinical-radiological model (AUC, 0.744; 95% CI, 0.642 - 0.846) in the training cohort. A nomogram based on the combined model achieved good clinical utility in predicting the treatment efficacy of TACE.

Conclusion

CE-MRI radiomics analysis may serve as a promising and noninvasive tool to predict therapeutic response to TACE in HCC, which will facilitate the individualized follow-up and further therapeutic strategies guidance in HCC patients.

Surveillance of Fontan-associated liver disease: current standards and a proposal from the European Society of Paediatric Radiology Abdominal Task Force

Since Francis Fontan first introduced the eponymous technique, the Fontan procedure, this type of surgical palliation has allowed thousands of children affected by specific heart malformations to reach adulthood. Nevertheless, abdominal, thoracic, lymphatic and neurologic complications are the price that is paid by these patients. Our review focuses on Fontan-associated liver disease; the purpose is to summarize the current understanding of its physiopathology, the aim of follow-up and the specific radiologic follow-up performed in Europe. Finally, we as members of the Abdominal Task Force of the European Society of Paediatric Radiology propose a consensus-based imaging follow-up algorithm.

Radiomics Analysis Based on Multiparametric MRI for Predicting Early Recurrence in Hepatocellular Carcinoma After Partial Hepatectomy

BACKGROUND

Preoperative prediction of early recurrence (ER) of hepatocellular carcinoma (HCC) plays a critical role in individualized risk stratification and further treatment guidance.

PURPOSE

To investigate the role of radiomics analysis based on multiparametric MRI (mpMRI) for predicting ER in HCC after partial hepatectomy.

STUDY TYPE

Retrospective.

POPULATION

In all, 113 HCC patients (ER, n = 58 vs. non-ER, n = 55), divided into training (n = 78) and validation (n = 35) cohorts.

FIELD STRENGTH/SEQUENCE

1.5T or 3.0T, gradient-recalled-echo in-phase T₁ -weighted imaging (I-T₁ WI) and opposed-phase T₁ WI (O-T₁ WI), fast spin-echo T₂ -weighted imaging (T₂ WI), spin-echo planar diffusion-weighted imaging (DWI), and gradient-recalled-echo contrast-enhanced MRI (CE-MRI).

ASSESSMENT

In all, 1146 radiomics features were extracted from each image sequence, and radiomics models based on each sequence and their combination were established via multivariate logistic regression analysis. The clinicopathologic-radiologic (CPR) model and the combined model integrating the radiomics score with the CPR risk factors were constructed. A nomogram based on the combined model was established.

STATISTICAL TESTS

Receiver operating characteristic (ROC) curve analysis was used to evaluate the discriminative performance of each model. The potential clinical usefulness was evaluated by decision curve analysis (DCA).

RESULTS

The radiomics model based on I-T₁ WI, O-T₁ WI, T₂ WI, and CE-MRI sequences presented the best performance among all radiomics models with an area under the ROC curve (AUC) of 0.771 (95% confidence interval (CI): 0.598-0.894) in the validation cohort. The combined nomogram (AUC: 0.873; 95% CI: 0.756-0.989) outperformed the radiomics model and the CPR model (AUC: 0.742; 95% CI: 0.577-0.907). DCA demonstrated that the combined nomogram was clinically useful.

DATA CONCLUSION

The mpMRI-based radiomics analysis has potential to predict ER of HCC patients after hepatectomy, which could enhance risk stratification and provide support for individualized treatment planning.

EVIDENCE LEVEL

4.

TECHNICAL EFFICACY

Stage 4.

Dynamic Hepatocellular Carcinoma Model Within a Liver Phantom for Multimodality Imaging

Introduction

Hepatocellular carcinoma (HCC) is one of the most common cancer in the world, and the effectiveness of its treatment lies in its detection in its early stages. The aim of this study is to mimic HCC dynamically through a liver phantom and apply it in multimodality medical imaging techniques including magnetic resonance imaging (MRI), computed tomography (CT), and ultrasound.

Methods and materials

The phantom is fabricated with two main parts, liver parenchyma and HCC inserts. The liver parenchyma was fabricated by adding 2.5 wt% of agarose powder combined with 2.6 wt% of wax powder while the basic material for the HCC samples was made from polyurethane solution combined with 5 wt% glycerol. Three HCC samples were inserted into the parenchyma by using three cylinders implanted inside the liver parenchyma. An automatic injector is attached to the input side of the cylinders and a suction device connected to the output side of the cylinders. After the phantom was prepared, the contrast materials were injected into the phantom and imaged using MRI, CT, and ultrasound.

Results

Both HCC samples and liver parenchyma were clearly distinguished using the three imaging modalities: MRI, CT, and ultrasound. Doppler ultrasound was also applied through the HCC samples and the flow pattern was observed through the samples.

Conclusion

A multimodal dynamic liver phantom, with HCC tumor models have been fabricated. This phantom helps to improve and develop different methods for detecting HCC in its early stages.

Radiomics in liver diseases: Current progress and future opportunities

Liver diseases, a wide spectrum of pathologies from inflammation to neoplasm, have become an increasingly significant health problem worldwide. Noninvasive imaging plays a critical role in the clinical workflow of liver diseases, but conventional imaging assessment may provide limited information. Accurate detection, characterization and monitoring remain challenging. With progress in quantitative imaging analysis techniques, radiomics emerged as an efficient tool that shows promise to aid in personalized diagnosis and treatment decision-making. Radiomics could reflect the heterogeneity of liver lesions via extracting high-throughput and high-dimensional features from multi-modality imaging. Machine learning algorithms are then used to construct clinical target-oriented imaging biomarkers to assist disease management. Here, we review the methodological process in liver disease radiomics studies in a stepwise fashion from data acquisition and curation, region of interest segmentation, liver-specific feature extraction, to task-oriented modelling. Furthermore, the applications of radiomics in liver diseases are outlined in aspects of diagnosis and staging, evaluation of liver tumour biological behaviours, and prognosis according to different disease type. Finally, we discuss the current limitations of radiomics in liver disease studies and explore its future opportunities.

Gadoxetic acid magnetic-enhanced resonance imaging in the diagnosis of cholangiocarcinoma

The use of liver magnetic resonance imaging is increasing thanks to its multiparametric sequences that allow a better tissue characterization, and the use of hepatobiliary contrast agents. This review aims to evaluate gadoxetic acid enhanced magnetic resonance imaging in the diagnosis and staging of cholangiocarcinoma and its different clinical and radiological classifications proposed in the literature. We also analyze the epidemiology, risk factors in correlation with clinical findings and laboratory data.

Hepatic tumors: pitfall in diagnostic imaging

On computed tomography (CT) and magnetic resonance imaging (MRI), hepatocellular tumors are characterized based on typical imaging findings. However, hepatocellular adenoma, focal nodular hyperplasia, and hepatocellular carcinoma can show uncommon appearances at CT and MRI, which may lead to diagnostic challenges. When assessing focal hepatic lesions, radiologists need to be aware of these atypical imaging findings to avoid misdiagnoses that can alter the management plan. The purpose of this review is to illustrate a variety of pitfalls and atypical features of hepatocellular tumors that can lead to misinterpretations providing specific clues to the correct diagnoses. (www.actabiomedica.it)

Role of dynamic perfusion magnetic resonance imaging in patients with local advanced rectal cancer

BACKGROUND

The management of rectal cancer patients is mainly based on the use of the magnetic resonance imaging (MRI) technique as a diagnostic tool for both staging and restaging. After treatment, to date, the evaluation of complete response is based on the histopathology assessment by using different tumor regression grade (TRG) features (e.g., Dworak or Mandard classifications). While from the radiological point of view, the main attention for the prediction of a complete response after chemotherapy treatment focuses on MRI and the potential role of diffusion-weighted images and perfusion imaging represented by dynamic-contrast enhanced MRI. The main aim is to find a reliable tool to predict tumor response in comparison to histopathologic findings.

AIM

To investigate the value of dynamic contrast-enhanced perfusion-MRI parameters in the evaluation of the healthy rectal wall and tumor response to chemo-radiation therapy in patients with local advanced rectal cancer with histopathologic correlation.

METHODS

Twenty-eight patients with biopsy-proven rectal adenocarcinoma who underwent a dynamic contrast-enhanced MR study performed on a 1.5T MRI system (Achieva, Philips), before (MR1) and after chemoradiation therapy (MR2), were enrolled in this study. The protocol included T1 gadolinium enhanced THRIVE sequences acquired on axial planes. A dedicated workstation was used to generate color permeability maps. Region of interest was manually drawn on tumor tissue and normal rectal wall, hence the following parameters were calculated and statistically analyzed: Relative arterial enhancement (RAE), relative venous enhancement (RVE), relative late enhancement (RLE), maximum enhancement (ME), time to peak and area under the curve (AUC). Perfusion parameters were related to pathologic TRG (Mandard's criteria; TRG1 = complete regression, TRG5 = no regression).

RESULTS

Ten tumors (36%) showed complete or subtotal regression (TRG1-2) at histology and classified as responders; 18 tumors (64%) were classified as non-responders (TRG3-5). Perfusion MRI parameters were significantly higher in the tumor tissue than in the healthy tissue in MR1 (P < 0.05). At baseline (MR1), no significant difference in perfusion parameters was found between responders and non-responders. After chemo-radiation therapy, at MR2, responders showed significantly (P < 0.05) lower perfusion values [RAE (%) 54 ± 20; RVE (%) 73 ± 24; RLE (%): 82 ± 29; ME (%): 904 ± 429] compared to non-responders [RAE (%): 129 ± 45; RVE (%): 154 ± 39; RLE (%): 164 ± 35; ME (%): 1714 ± 427]. Moreover, in responders group perfusion values decreased significantly at MR2 [RAE (%): 54 ± 20; RVE (%): 73 ± 24; RLE (%): 82 ± 29; ME (%): 904 ± 429] compared to the corresponding perfusion values at MR1 [RAE (%): 115 ± 21; RVE (%): 119 ± 21; RLE (%): 111 ± 74; ME (%): 1060 ± 325]; (P < 0.05). Concerning the time-intensity curves, the AUC at MR2 showed significant difference (P = 0.03) between responders and non-responders [AUC (mm² × 10^-3) 121 ± 50 vs 258 ± 86], with lower AUC values of the tumor tissue in responders compared to non-responders. In non-responders, there were no significant differences between perfusion values at MR1 and MR2.

CONCLUSION

Dynamic contrast perfusion-MRI analysis represents a complementary diagnostic tool for identifying vascularity characteristics of tumor tissue in local advanced rectal cancer, useful in the assessment of treatment response.

Relationships and Qualitative Evaluation Between Diffusion-Weighted Imaging and Pathologic Findings of Resected Lung Cancers

For detecting malignant tumors, diffusion-weighted magnetic resonance imaging (DWI) as well as fluoro-2-deoxy-glucose positron emission tomography/computed tomography (FDG-PET/CT) are available. It is not definitive how DWI correlates the pathological findings of lung cancer. The aim of this study is to evaluate the relationships between DWI findings and pathologic findings. In this study, 226 patients with resected lung cancers were enrolled. DWI was performed on each patient before surgery. There were 167 patients with adenocarcinoma, 44 patients with squamous cell carcinoma, and 15 patients with other cell types. Relationships between the apparent diffusion coefficient (ADC) of DWI and the pathology were analyzed. When the optimal cutoff value (OCV) of ADC for diagnosing malignancy was 1.70 × 10⁻³ mm²/s, the sensitivity of DWI was 92.0% (208/226). The sensitivity was 33.3% (3/9) in mucinous adenocarcinoma. The ADC value (1.31 ± 0.32 × 10⁻³ mm²/s) of adenocarcinoma was significantly higher than that (1.17 ± 0.29 × 10⁻³ mm²/s) of squamous cell carcinoma (p = 0.012), or (0.93 ± 0.14 × 10⁻³ mm²/s) of small cell carcinoma (p = 0.0095). The ADC value (1.91 ± 0.36 × 10⁻³ mm²/s) of mucinous adenocarcinoma was significantly higher than that (1.25 ± 0.25 × 10⁻³ mm²/s) of adenocarcinoma with mucin and that (1.24 ± 0.30 × 10⁻³ mm²/s) of other cell types. The ADC (1.11 ± 0.26 × 10⁻³ mm²/s) of lung cancer with necrosis was significantly lower than that (1.32 ± 0.33 × 10⁻³ mm²/s) of lung cancer without necrosis. The ADC of mucinous adenocarcinoma was significantly higher than those of adenocarcinoma of other cell types. The ADC of lung cancer was likely to decrease according to cell differentiation decreasing. The sensitivity of DWI for lung cancer was 92% and this result shows that DWI is valuable for the evaluation of lung cancer. Lung cancer could be evaluated qualitatively using DWI.

Characterization of the arterial enhancement pattern of focal liver lesions by multiple arterial phase magnetic resonance imaging: comparison between hepatocellular carcinoma and focal nodular hyperplasia

PURPOSE

To evaluate the features of arterial enhancement pattern of focal nodular hyperplasia (FNH) and hepatocellular carcinoma (HCC) by triple-phase arterial magnetic resonance imaging (MRI).

METHODS

Data were retrospectively collected from 52 consecutive patients who underwent triple-phase arterial MRI using hepatocyte-specific contrast agents (Gd-EOB-DTPA) from January 2017 to October 2017, with a MR imaging diagnosis of HCC or FNH. The images were independently assessed by two blinded readers. Contrast enhancement ratio (CER) and liver-to-lesion contrast ratio (LLCR) were calculated. The lesions were classified visually and also based on the peak of LLCR into the following groups: (1) early arterial, (2) middle arterial and (3) late arterial. Data were eventually analysed using nonparametric tests.

RESULTS

The CER analysis showed no significant difference between HCC and FNH patients (p > 0.05). LLCR_FNH were significantly higher than LLCR_HCC in the early arterial (p = 0.01), but not in the middle and late arterial phases (p = 0.20 and p = 0.82, respectively). LLCR_HCC presented a meaningful increase from early to middle arterial phase (p = 0.009), whereas LLCR_FNH showed a decrease from middle to late arterial phase (p = 0.004). Based on the peak of LLCR, 17 (55%) FNHs were classified into early, 11 (35%) in middle and only 3 (10%) in late arterial phase groups. Similarly, 14 (34%) HCCs were categorized into early, 13 (32%) in middle and 14 (33%) in late arterial phase groups. There was a good agreement between qualitative analysis and LLCR in 85% of cases.

CONCLUSION

The optimal visualization of FNH has been detected in early and middle arterial phases while HCC has been best observed during middle and late arterial phases.

Application Value of Magnetic Resonance Perfusion Imaging in the Early Diagnosis of Rat Hepatic Fibrosis

Objective

To assess the application value of perfusion-weighted imaging (PWI) in early diagnosis, quantitation, and hepatic fibrosis staging by analyzing the related parameters in hepatic fibrosis.

Methods

A total of 60 rats were randomly divided into the hepatic fibrosis and control groups, and carbon tetrachloride (CCL4) was used to establish the liver fibrosis model. All rats underwent PWI examination, and the trend of the time-signal intensity curve (TIC, automatically generated by the software) was observed. Also, the perfusion parameters, maximum signal reduction ratio (SRRmax), time to peak (TTP), and mean transit time (MTT), were analyzed and compared with pathological staging.

Results

The TIC curve was characterized by slow wash-in and wash-out with a low and wide peak. The PWI perfusion parameters were statistically significant in specific groups (P < 0.05): SRRmax values (control group and F3, F4), TTP, and MTT values (control group and F2–F4, F1 and F3, F1 and F4, and F2 and F4 in addition to TTP values for F1 and F2). Pearson's correlation analysis showed a negative correlation of SRRmax with hepatic fibrosis stage (r = −0.439, P < 0.05), while TTP and MTT values were positively correlated with hepatic fibrosis stage (TTP, r = 0.798; MTT, r = 0.647; all P < 0.001).

Conclusions

PWI perfusion parameters reflect the degree of hepatic fibrosis, especially TTP and MTT, and PWI is recommended for the early diagnosis of liver fibrosis for timely intervention and treatment of the disease and delaying its progression.

Biomarkers vs imaging in the early detection of hepatocellular carcinoma and prognosis

Hepatocellular carcinoma (HCC) is the 5^th most frequently diagnosed cancer in the world, according to the World Health Organization. The incidence of HCC is between 3/100000 and 78.1/100000, with a high incidence reported in areas with viral hepatitis B and hepatitis C, thus affecting Asia and Africa predominantly. Several international clinical guidelines address HCC diagnosis and are structured according to the geographical area involved. All of these clinical guidelines, however, share a foundation of diagnosis by ultrasound surveillance and contrast imaging techniques, particularly computed tomography, magnetic resonance imaging, and sometimes contrast-enhanced ultrasound. The primary objective of this review was to systematically summarize the recent published studies on the clinical utility of serum biomarkers in the early diagnosis of HCC and for the prognosis of this disease.

Preoperative diagnosis and prediction of hepatocellular carcinoma: Radiomics analysis based on multi-modal ultrasound images

BACKGROUND

This study aims to establish a radiomics analysis system for the diagnosis and clinical behaviour prediction of hepatocellular carcinoma (HCC) based on multi-parametric ultrasound imaging.

METHODS

A total of 177 patients with focal liver lesions (FLLs) were included in the study. Every patient underwent multi-modal ultrasound examination, including B-mode ultrasound (BMUS), shear wave elastography (SWE), and shear wave viscosity (SWV) imaging. The radiomics analysis system was built on sparse representation theory (SRT) and support vector machine (SVM) for asymmetric data. Through the sparse regulation from the SRT, the proposed radiomics system can effectively avoid over-fitting issues that occur in regular radiomics analysis. The purpose of the proposed system includes differential diagnosis between benign and malignant FLLs, pathologic diagnosis of HCC, and clinical prognostic prediction. Three biomarkers, including programmed cell death protein 1 (PD-1), antigen Ki-67 (Ki-67) and microvascular invasion (MVI), were included and analysed. We calculated the accuracy (ACC), sensitivity (SENS), specificity (SPEC) and area under the receiver operating characteristic curve (AUC) to evaluate the performance of the radiomics models.

RESULTS

A total of 2560 features were extracted from the multi-modal ultrasound images for each patient. Five radiomics models were built, and leave-one-out cross-validation (LOOCV) was used to evaluate the models. In LOOCV, the AUC was 0.94 for benign and malignant classification (95% confidence interval [CI]: 0.88 to 0.98), 0.97 for malignant subtyping (95% CI: 0.93 to 0.99), 0.97 for PD-1 prediction (95% CI: 0.89 to 0.98), 0.94 for Ki-67 prediction (95% CI: 0.87 to 0.97), and 0.98 for MVI prediction (95% CI: 0.93 to 0.99). The performance of each model improved when the viscosity modality was included.

CONCLUSIONS

Radiomics analysis based on multi-modal ultrasound images could aid in comprehensive liver tumor evaluations, including diagnosis, differential diagnosis, and clinical prognosis.