Machine Learning-Based Ultrasomics Improves the Diagnostic Performance in Differentiating Focal Nodular Hyperplasia and Atypical Hepatocellular Carcinoma

The Index sAGP is Valuable for Distinguishing Atypical Hepatocellular Carcinoma from Atypical Benign Focal Hepatic Lesions

Purpose

The differential diagnosis of atypical hepatocellular carcinoma (aHCC) and atypical benign focal hepatic lesions (aBFHL) usually depends on pathology. This study aimed to develop non-invasive approaches based on conventional blood indicators for the differential diagnosis of aHCC and aBFHL.

Patients and Methods

Hospitalized patients with pathologically confirmed focal hepatic lesions and their clinical data were retrospectively collected, in which patients with HCC with serum alpha-fetoprotein (AFP) levels of ≤200 ng/mL and atypical imaging features were designated as the aHCC group (n = 224), and patients with benign focal hepatic lesions without typical imaging features were designated as the aBFHL group (n = 178). The performance of indexes (both previously reported and newly constructed) derived from conventional blood indicators by four mathematical operations in distinguishing aHCC and aBFHL was evaluated using the receiver operating characteristic (ROC) curve and diagnostic validity metrics.

Results

Among ten previously reported derived indexes related to HCC, the index GPR, the ratio of γ-glutamyltransferase (GGT) to platelet (PLT), showed the best performance in distinguishing aHCC from aBFHL with the area under ROC curve (AUROC) of 0.853 (95% CI 0.814-0.892), but the other indexes were of little value (AUROCs from 0.531 to 0.700). A new derived index, sAGP [(standardized AFP + standardized GGT)/standardized PLT], was developed and exhibited AUROCs of 0.905, 0.894, 0.891, 0.925, and 0.862 in differentiating overall, BCLC stage 0/A, TNM stage I, small, and AFP-negative aHCC from aBFHL, respectively.

Conclusion

The sAGP index is an efficient, simple, and practical metric for the non-invasive differentiation of aHCC from aBFHL.

Perfluorobutane application value in microwave ablation of Small Hepatocellular Carcinoma (<3 cm)

BACKGROUND

No studies have been retrieved comparing perfluorobutane with sulfur hexafluoride for microwave ablation (MWA) in small hepatocellular carcinoma(sHCC).

OBJECTIVE

To retrospective investigate the value of perfluorobutane ultrasonography contrast agent in ultrasonography (US)-guided MWA of sHCC.

METHODS

We conducted a retrospective clinical controlled study about US-guided percutaneous MWA in patients with sHCC, and in patients undergoing intra-operative treatment with perfluorobutane or sulfur hexafluoride. In both groups, a contrast agent was injected to clear the tumor and then a needle was inserted. A 5-point needle prick difficulty score was developed to compare needle prick difficulty in the two groups of cases.

RESULTS

A total of 67 patients were included: 25 patients in group perfluorobutane, aged 41-82 (60.64±9.46), tumor size 1.1-2.8 (1.78±0.45) cm. 42 patients in group sulfur hexafluoride, aged 38-78 (62.26±9.27), with tumor size of 1.1-3.0 (1.89±0.49) cm. There was no significant difference in age or tumor size in both groups (P >  0.05). Puncture difficulty score (5-point): 2.0-2.7 (2.28±0.29) in group perfluorobutane, and 2.0-4.7 (2.95±0.85) in group sulfur hexafluoride, and the difference between the two groups was statistically significant (P <  0.05). Enhanced imaging results within 3 months after surgery: complete ablation rate was 100% (25/25) in the group perfluorobutane, 95.2% (40/42 in the group sulfur hexafluoride), with no significant difference between the two groups (P >  0.05).

CONCLUSION

Perfluorobutane kupffer phase can make the operator accurately deploy the ablation needle and reduce the difficulty of operation.

Deep learning approach for discrimination of liver lesions using nine time-phase images of contrast-enhanced ultrasound

PURPOSE

Contrast-enhanced ultrasound (CEUS) shows different enhancement patterns depending on the time after administration of the contrast agent. The aim of this study was to evaluate the diagnostic performance of liver nodule characterization using our proposed deep learning model with input of nine CEUS images.

METHODS

A total of 181 liver lesions (48 benign, 78 hepatocellular carcinoma (HCC), and 55 non-HCC malignant) were included in this prospective study. CEUS were performed using the contrast agent Sonazoid, and in addition to B-mode images before injection, image clips were stored every minute up to 10 min. A deep learning model was developed by arranging three ResNet50 transfer learning models in parallel. This proposed model allowed inputting up to nine datasets of different phases of CEUS and performing image augmentation of nine images synchronously. Using the results, the correct prediction rate, sensitivity, and specificity between "benign" and "malignant" cases were analyzed for each combination of the time phase. These accuracy values were also compared with the washout score judged by a human.

RESULTS

The proposed model showed performance superior to the referential standard model when the dataset from B-mode to the 10-min images were used (sensitivity: 93.2%, specificity: 65.3%, average correct answer rate: 60.1%). It also maintained 90.2% sensitivity and 61.2% specificity even when the dataset was limited to 2 min after injection, and this accuracy was equivalent to or better than human scoring by experts.

CONCLUSION

Our proposed model has the potential to identify tumor types earlier than the Kupffer phase, but at the same time, machine learning confirmed that Kupffer-phase Sonazoid images contain essential information for the classification of liver nodules.

Diagnostic Performance of an Artificial Intelligence Model Based on Contrast-Enhanced Ultrasound in Patients with Liver Lesions: A Comparative Study with Clinicians

Contrast-enhanced ultrasound (CEUS) is widely used in the characterization of liver tumors; however, the evaluation of perfusion patterns using CEUS has a subjective character. This study aims to evaluate the accuracy of an automated method based on CEUS for classifying liver lesions and to compare its performance with that of two experienced clinicians. The system used for automatic classification is based on artificial intelligence (AI) algorithms. For an interpretation close to the clinical setting, both clinicians knew which patients were at high risk for hepatocellular carcinoma (HCC), but only one was aware of all the clinical data. In total, 49 patients with 59 liver tumors were included. For the benign and malignant classification, the AI model outperformed both clinicians in terms of specificity (100% vs. 93.33%); still, the sensitivity was lower (74% vs. 93.18% vs. 90.91%). In the second stage of multiclass diagnosis, the automatic model achieved a diagnostic accuracy of 69.93% for HCC and 89.15% for liver metastases. Readers demonstrated greater diagnostic accuracy for HCC (83.05% and 79.66%) and liver metastases (94.92% and 96.61%) compared to the AI system; however, both were experienced sonographers. The AI model could potentially assist and guide less-experienced clinicians to discriminate malignant from benign liver tumors with high accuracy and specificity.

Current Status and Analysis of Machine Learning in Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is a common tumor. Although the diagnosis and treatment of HCC have made great progress, the overall prognosis remains poor. As the core component of artificial intelligence, machine learning (ML) has developed rapidly in the past decade. In particular, ML has become widely used in the medical field, and it has helped in the diagnosis and treatment of cancer. Different algorithms of ML have different roles in diagnosis, treatment, and prognosis. This article reviews recent research, explains the application of different ML models in HCC, and provides suggestions for follow-up research.

Machine learning for malignant versus benign focal liver lesions on US and CEUS: a meta-analysis

OBJECTIVES

To perform a meta-analysis of the diagnostic performance of learning (ML) algorithms (conventional and deep learning algorithms) for the classification of malignant versus benign focal liver lesions (FLLs) on US and CEUS.

METHODS

Available databases were searched for relevant published studies through September 2022. Studies met eligibility criteria if they evaluate the diagnostic performance of ML for the classification of malignant and benign focal liver lesions on US and CEUS. The pooled per-lesion sensitivities and specificities for each modality with 95% confidence intervals were calculated.

RESULTS

A total of 8 studies on US, 11 on CEUS, and 1 study evaluating both methods met the inclusion criteria with a total of 34,245 FLLs evaluated. The pooled sensitivity and specificity of ML for the malignancy classification of FLLs were 81.7% (95% CI, 77.2-85.4%) and 84.8% (95% CI, 76.0-90.8%) for US, compared to 87.1% (95% CI, 81.8-91.0%) and 87.0% (95% CI, 83.1-90.1%) for CEUS. In the subgroup analysis of studies that evaluated deep learning algorithms, the sensitivity and specificity of CEUS (n = 4) increased to 92.4% (95% CI, 88.5-95.0%) and 88.2% (95% CI, 81.1-92.9%).

CONCLUSIONS

The diagnostic performance of ML algorithms for the malignant classification of FLLs was high for both US and CEUS with overall similar sensitivity and specificity. The similar performance of US may be related to the higher prevalence of DL models in that group.

Artificial intelligence for the classification of focal liver lesions in ultrasound - a systematic review

Focal liver lesions are detected in about 15% of abdominal ultrasound examinations. The diagnosis of frequent benign lesions can be determined reliably based on the characteristic B-mode appearance of cysts, hemangiomas, or typical focal fatty changes. In the case of focal liver lesions which remain unclear on B-mode ultrasound, contrast-enhanced ultrasound (CEUS) increases diagnostic accuracy for the distinction between benign and malignant liver lesions. Artificial intelligence describes applications that try to emulate human intelligence, at least in subfields such as the classification of images. Since ultrasound is considered to be a particularly examiner-dependent technique, the application of artificial intelligence could be an interesting approach for an objective and accurate diagnosis. In this systematic review we analyzed how artificial intelligence can be used to classify the benign or malignant nature and entity of focal liver lesions on the basis of B-mode or CEUS data. In a structured search on Scopus, Web of Science, PubMed, and IEEE, we found 52 studies that met the inclusion criteria. Studies showed good diagnostic performance for both the classification as benign or malignant and the differentiation of individual tumor entities. The results could be improved by inclusion of clinical parameters and were comparable to those of experienced investigators in terms of diagnostic accuracy. However, due to the limited spectrum of lesions included in the studies and a lack of independent validation cohorts, the transfer of the results into clinical practice is limited.

Artificial intelligence models for the diagnosis and management of liver diseases

With the development of more advanced methods for the diagnosis and treatment of diseases, the data required for medical care are becoming complex, and misinterpretation of information due to human error may result in serious consequences. Human error can be avoided with the support of artificial intelligence (AI). AI models trained with various medical data for diagnosis and management of liver diseases have been applied to hepatitis, fatty liver disease, liver cirrhosis, and liver cancer. Some of these models have been reported to outperform human experts in terms of performance, indicating their potential for supporting clinical practice given their high-speed output. This paper summarizes the recent advances in AI for liver disease and introduces the AI-aided diagnosis of liver tumors using B-mode ultrasonography.

Serological indices and ultrasound variables in predicting the staging of hepatitis B liver fibrosis: A comparative study based on random forest algorithm and traditional methods

Objective

To compare the diagnostic efficacy of serological indices and ultrasound (US) variables in hepatitis B virus (HBV) liver fibrosis staging using random forest algorithm (RFA) and traditional methods.

Methods

The demographic and serological indices and US variables of patients with HBV liver fibrosis were retrospectively collected and divided into serology group, US group, and serology + US group according to the research content. RFA was used for training and validation. The diagnostic efficacy was compared to logistic regression analysis (LRA) and APRI and FIB-4 indices.

Results

For the serology group, the diagnostic performance of RFA was significantly higher than that of APRI and FIB-4 indices. The diagnostic accuracy of RFA in the four classifications (S0S1/S2/S3/S4) of the hepatic fibrosis stage was 79.17%. The diagnostic accuracy for significant fibrosis (≥S2), advanced fibrosis (≥S3), and cirrhosis (S4) was 87.99%, 90.69%, and 92.40%, respectively. The area under the curve (AUC) values were 0.945, 0.959, and 0.951, respectively. For the US group, there was no significant difference in diagnostic performance between RFA and LRA. The diagnostic performance of RFA in the serology + US group was significantly better than that of LRA. The diagnostic accuracy of the four classifications (S0S1/S2/S3/S4) of the hepatic fibrosis stage was 77.21%. The diagnostic accuracy for significant fibrosis (≥S2), advanced fibrosis (≥S3), and cirrhosis (S4) was 87.50%, 90.93%, and 93.38%, respectively. The AUC values were 0.948, 0.959, and 0.962, respectively.

Conclusion

RFA can significantly improve the diagnostic performance of HBV liver fibrosis staging. RFA based on serological indices has a good ability to predict liver fibrosis staging. RFA can help clinicians accurately judge liver fibrosis staging and reduce unnecessary biopsies.

Multimodal ultrasound fusion network for differentiating between benign and malignant solid renal tumors

Objective: We aim to establish a deep learning model called multimodal ultrasound fusion network (MUF-Net) based on gray-scale and contrast-enhanced ultrasound (CEUS) images for classifying benign and malignant solid renal tumors automatically and to compare the model’s performance with the assessments by radiologists with different levels of experience.

Methods: A retrospective study included the CEUS videos of 181 patients with solid renal tumors (81 benign and 100 malignant tumors) from June 2012 to June 2021. A total of 9794 B-mode and CEUS-mode images were cropped from the CEUS videos. The MUF-Net was proposed to combine gray-scale and CEUS images to differentiate benign and malignant solid renal tumors. In this network, two independent branches were designed to extract features from each of the two modalities, and the features were fused using adaptive weights. Finally, the network output a classification score based on the fused features. The model’s performance was evaluated using five-fold cross-validation and compared with the assessments of the two groups of radiologists with different levels of experience.

Results: For the discrimination between benign and malignant solid renal tumors, the junior radiologist group, senior radiologist group, and MUF-Net achieved accuracy of 70.6%, 75.7%, and 80.0%, sensitivity of 89.3%, 95.9%, and 80.4%, specificity of 58.7%, 62.9%, and 79.1%, and area under the receiver operating characteristic curve of 0.740 (95% confidence internal (CI): 0.70–0.75), 0.794 (95% CI: 0.72–0.83), and 0.877 (95% CI: 0.83–0.93), respectively.

Conclusion: The MUF-Net model can accurately classify benign and malignant solid renal tumors and achieve better performance than senior radiologists.

Key points: The CEUS video data contain the entire tumor microcirculation perfusion characteristics. The proposed MUF-Net based on B-mode and CEUS-mode images can accurately distinguish between benign and malignant solid renal tumors with an area under the receiver operating characteristic curve of 0.877, which surpasses senior radiologists’ assessments by a large margin.

Deep Learning for Approaching Hepatocellular Carcinoma Ultrasound Screening Dilemma: Identification of α-Fetoprotein-Negative Hepatocellular Carcinoma From Focal Liver Lesion Found in High-Risk Patients

Background

First-line surveillance on hepatitis B virus (HBV)-infected populations with B-mode ultrasound is relatively limited to identifying hepatocellular carcinoma (HCC) without elevated α-fetoprotein (AFP). To improve the present HCC surveillance strategy, the state of the art of artificial intelligence (AI), a deep learning (DL) approach, is proposed to assist in the diagnosis of a focal liver lesion (FLL) in HBV-infected liver background.

Methods

Our proposed deep learning model was based on B-mode ultrasound images of surgery that proved 209 HCC and 198 focal nodular hyperplasia (FNH) cases with 413 lesions. The model cohort and test cohort were set at a ratio of 3:1, in which the test cohort was composed of AFP-negative HBV-infected cases. Four additional deep learning models (MobileNet, Resnet50, DenseNet121, and InceptionV3) were also constructed as comparative baselines. To evaluate the models in terms of diagnostic power, sensitivity, specificity, accuracy, confusion matrix, F1-score, and area under the receiver operating characteristic curve (AUC) were calculated in the test cohort.

Results

The AUC of our model, Xception, achieved 93.68% in the test cohort, superior to other baselines (89.06%, 85.67%, 83.94%, and 78.13% respectively for MobileNet, Resnet50, DenseNet121, and InceptionV3). In terms of diagnostic power, our model showed sensitivity, specificity, accuracy, and F1-score of 96.08%, 76.92%, 86.41%, and 87.50%, respectively, and PPV, NPV, FPR, and FNR calculated from the confusion matrix were respectively 80.33%, 95.24%, 23.08%, and 3.92% in identifying AFP-negative HCC from HBV-infected FLL cases. Satisfactory robustness of our proposed model was shown based on 5-fold cross-validation performed among the models above.

Conclusions

Our DL approach has great potential to assist B-mode ultrasound in identifying AFP-negative HCC from FLL found in surveillance of HBV-infected patients.

Recent Advances in Machine Learning Applied to Ultrasound Imaging

Machine learning (ML) methods are pervading an increasing number of fields of application because of their capacity to effectively solve a wide variety of challenging problems. The employment of ML techniques in ultrasound imaging applications started several years ago but the scientific interest in this issue has increased exponentially in the last few years. The present work reviews the most recent (2019 onwards) implementations of machine learning techniques for two of the most popular ultrasound imaging fields, medical diagnostics and non-destructive evaluation. The former, which covers the major part of the review, was analyzed by classifying studies according to the human organ investigated and the methodology (e.g., detection, segmentation, and/or classification) adopted, while for the latter, some solutions to the detection/classification of material defects or particular patterns are reported. Finally, the main merits of machine learning that emerged from the study analysis are summarized and discussed.

Development and Validation of a Radiomic Nomogram for Predicting the Prognosis of Kidney Renal Clear Cell Carcinoma

Purpose

The present study aims to comprehensively investigate the prognostic value of a radiomic nomogram that integrates contrast-enhanced computed tomography (CECT) radiomic signature and clinicopathological parameters in kidney renal clear cell carcinoma (KIRC).

Methods

A total of 136 and 78 KIRC patients from the training and validation cohorts were included in the retrospective study. The intraclass correlation coefficient (ICC) was used to assess reproducibility of radiomic feature extraction. Univariate Cox analysis and least absolute shrinkage and selection operator (LASSO) as well as multivariate Cox analysis were utilized to construct radiomic signature and clinical signature in the training cohort. A prognostic nomogram was established containing a radiomic signature and clinicopathological parameters by using a multivariate Cox analysis. The predictive ability of the nomogram [relative operating characteristic curve (ROC), concordance index (C-index), Hosmer–Lemeshow test, and calibration curve] was evaluated in the training cohort and validated in the validation cohort. Patients were split into high- and low-risk groups, and the Kaplan–Meier (KM) method was conducted to identify the forecasting ability of the established models. In addition, genes related with the radiomic risk score were determined by weighted correlation network analysis (WGCNA) and were used to conduct functional analysis.

Results

A total of 2,944 radiomic features were acquired from the tumor volumes of interest (VOIs) of CECT images. The radiomic signature, including ten selected features, and the clinical signature, including three selected clinical variables, showed good performance in the training and validation cohorts [area under the curve (AUC), 0.897 and 0.712 for the radiomic signature; 0.827 and 0.822 for the clinical signature, respectively]. The radiomic prognostic nomogram showed favorable performance and calibration in the training cohort (AUC, 0.896, C-index, 0.846), which was verified in the validation cohort (AUC, 0.768). KM curves indicated that the progression-free interval (PFI) time was dramatically shorter in the high-risk group than in the low-risk group. The functional analysis indicated that radiomic signature was significantly associated with T cell activation.

Conclusions

The nomogram combined with CECT radiomic and clinicopathological signatures exhibits excellent power in predicting the PFI of KIRC patients, which may aid in clinical management and prognostic evaluation of cancer patients.

Distinguishing Hepatocellular Carcinoma From Hepatic Inflammatory Pseudotumor Using a Nomogram Based on Contrast-Enhanced Ultrasound

BACKGROUND

Hepatocellular carcinoma (HCC) and hepatic iflammatory pseudotumor (IPT) share similar symptoms and imaging features, which makes it challenging to distinguish from each other in clinical practice. This study aims to develop a predictive model based on contrast-enhanced ultrasound (CEUS) and clinical features to discriminate HCC from IPT.

METHODS

Sixty-two IPT and 146 HCC patients were enrolled in this study, where pathological diagnosis served as the reference standard for diagnosis. Clinical and ultrasound imaging data including CEUS features: enhancement degree during arterial phase, portal phase and delayed phase, enhancement pattern, early washout within 60 s, feeding artery, peritumoral vessels, peritumoral enhancement, and margin of nonenhanced area were retrospectively collected. Imaging data were reviewed by two experienced ultrasound doctors. Patients were randomly assigned to training and validation sets. Chi-squared test followed by LASSO regression was performed on ultrasonographic features in the training set to identify the most valuable features that distinguish HCC from IPT, based on which the sonographic score formula was generated. With the significant clinical and ultrasonographic indicators, a nomogram was developed. The performance of the nomogram was verified by ROC curve and decision curve analysis (DCA) with the comparison with sonographic score and the ultrasound doctor's diagnosis.

RESULTS

The most valuable ultrasonographic features that distinguish between HCC and IPT were enhancement degree during arterial phase, early washout, peritumoral vessels, peritumoral enhancement, and liver background. The sonographic score based on these features was verified to be an independent factor that predicts the diagnosis (p = 0.003). Among the clinical indicators, AFP (p = 0.009) and viral hepatitis infection (p = 0.004) were significant. Sonographic score, AFP, and viral hepatitis were used to construct a predictive nomogram. The AUC of the nomogram was 0.989 and 0.984 in training and validation sets, respectively, which were higher than those of sonographic score alone (0.938 and 0.958) or the ultrasound doctor's diagnosis (0.794 and 0.832). DCA showed the nomogram provided the greatest clinical usefulness.

CONCLUSION

A predictive nomogram based on a sonographic signature improved the diagnostic performance in distinguishing HCC and IPT, which may help with individualized diagnosis and treatment in clinical practice.