Point Shear Wave Elastography Using Machine Learning to Differentiate Renal Cell Carcinoma and Angiomyolipoma

A narrative review on the application of artificial intelligence in renal ultrasound

Kidney disease is a serious public health problem and various kidney diseases could progress to end-stage renal disease. The many complications of end-stage renal disease. have a significant impact on the physical and mental health of patients. Ultrasound can be the test of choice for evaluating the kidney and perirenal tissue as it is real-time, available and non-radioactive. To overcome substantial interobserver variability in renal ultrasound interpretation, artificial intelligence (AI) has the potential to be a new method to help radiologists make clinical decisions. This review introduces the applications of AI in renal ultrasound, including automatic segmentation of the kidney, measurement of the renal volume, prediction of the kidney function, diagnosis of the kidney diseases. The advantages and disadvantages of the applications will also be presented clinicians to conduct research. Additionally, the challenges and future perspectives of AI are discussed.

Alkaline phosphatase combines with CT factors for differentiating small (≤ 4 cm) fat-poor angiomyolipoma from renal cell carcinoma: a multiple quantitative tool

PURPOSE

This study aimed to evaluate the diagnostic value of serum and CT factors to establish a convenient diagnostic method for differentiating small (≤ 4 cm) fat-poor angiomyolipoma (AML) from renal cell carcinoma (RCC).

MATERIALS AND METHODS

This study analyzed the preoperative serum laboratory data and CT data of 32 fat-poor AML patients and 133 RCC patients. The CT attenuation value of tumor (AVT), relative enhancement ratio (RER), and heterogeneous degree of tumor were detected using region of interest on precontrast phase (PCP) and the corticomedullary phase. Multivariate regression was performed to filter the main factors. The main factors were selected to establish the prediction models. The area under the curve (AUC) was measured to evaluate the diagnostic efficacy.

RESULTS

Fat-poor AML was more common found in younger (47.91 ± 2.09 years vs 53.63 ± 1.17 years, P = 0.02) and female (70.68 vs 28.13%, P < 0.001) patients. Alkaline phosphatase (ALP) was higher in RCC patients (81.80 ± 1.75 vs 63.25 ± 2.95 U/L, P < 0.01). For CT factors, fat-poor AML was higher in PCP_AVT (40.30 ± 1.49 vs 32.98 ± 0.69Hu, P < 0.01) but lower in RER (67.17 ± 3.17 vs 84.64 ± 2.73, P < 0.01). Gender, ALP, PCP_AVT and RER was found valuable for the differentiation. When compared with laboratory-based or CT-based diagnostic models, the combination model integrating gender, ALP, PCP_AVT and RER shows the best diagnostic performance (AUC = 0.922).

CONCLUSION

ALP was found higher in RCC patients. Female patients with ALP < 70.50U/L, PCP_AVT > 35.97Hu and RER < 82.66 are more likely to be diagnose as fat-poor AML.

Ultrasound Image-Based Deep Features and Radiomics for the Discrimination of Small Fat-Poor Angiomyolipoma and Small Renal Cell Carcinoma

We evaluated the performance of ultrasound image-based deep features and radiomics for differentiating small fat-poor angiomyolipoma (sfp-AML) from small renal cell carcinoma (SRCC). This retrospective study included 194 patients with pathologically proven small renal masses (diameter ≤4 cm; 67 in the sfp-AML group and 127 in the SRCC group). We obtained 206 and 364 images from the sfp-AML and SRCC groups with experienced radiologist identification, respectively. We extracted 4024 deep features from the autoencoder neural network and 1497 radiomics features from the Pyradiomics toolbox; the latter included first-order, shape, high-order, Laplacian of Gaussian and Wavelet features. All subjects were allocated to the training and testing sets with a ratio of 3:1 using stratified sampling. The least absolute shrinkage and selection operator (LASSO) regression model was applied to select the most diagnostic features. Support vector machine (SVM) was adopted as the discriminative classifier. An optimal feature subset including 45 deep and 7 radiomics features was screened by the LASSO model. The SVM classifier achieved good performance in discriminating between sfp-AMLs and SRCCs, with areas under the curve (AUCs) of 0.96 and 0.85 in the training and testing sets, respectively. The classifier built using deep and radiomics features can accurately differentiate sfp-AMLs from SRCCs on ultrasound imaging.

Elastography in the Urological Practice: Urinary and Male Genital Tract, Prostate Excluded—Review

The aim of this article is to review the utility of elastography in the day-to-day clinical practice of the urologist. An electronic database search was performed on PubMed and Cochrane Library with a date range between January 2000 and December 2021. The search yielded 94 articles that passed the inclusion and exclusion criteria. The articles were reviewed and discussed by organ, pathology and according to the physical principle underlying the elastographic method. Elastography was used in the study of normal organs, tumoral masses, chronic upper and lower urinary tract obstructive diseases, dysfunctions of the lower urinary tract and the male reproductive system, and as a pre- and post-treatment monitoring tool. Elastography has numerous applications in urology, but due to a lack of standardization in the methodology and equipment, further studies are required.

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.

Preoperative diagnosis and prediction of microvascular invasion in hepatocellularcarcinoma by ultrasound elastography

Background

To assess the values of two elastography techniques combined with serological examination and clinical features in preoperative diagnosis of microvascular invasion in HCC patients.

Methods

A total of 74 patients with single Hepatocellular carcinoma (HCC) were included in this study. Shear wave measurement and real-time tissue elastography were used to evaluate the hardness of tumor-adjacent tissues and tumor tissues, as well as the strain rate ratio per lesion before surgery. According to the pathological results, the ultrasound parameters and clinical laboratory indicators related to microvascular invasion were analyzed, and the effectiveness of each parameter in predicting the occurrence of microvascular invasion was compared.

Results

33/74 patients exhibited microvascular invasion. Univariate analysis showed that the hardness of tumor-adjacent tissues (P = 0.003), elastic strain rate ratio (P = 0.032), maximum tumor diameter (P < 0.001), and alpha-fetoprotein (AFP) level (P = 0.007) was significantly different in the patients with and without microvascular invasion. The binary logistic regression analysis showed that the maximum tumor diameter (P = 0.001) was an independent risk factor for predicting microvascular invasion, while the hardness of tumor-adjacent tissues (P = 0.028) was a protective factor. The receiver operating characteristic (ROC) curve showed that the area under the curve (AUC) of the hardness of tumor-adjacent tissues, the maximum diameter of the tumor, and the predictive model Logit(P) in predicting the occurrence of MVI was 0.718, 0.775 and 0.806, respectively.

Conclusion

The hardness of tumor-adjacent tissues, maximum tumor diameter, and the preoperative prediction model predict the occurrence of MVI in HCC patients.

Hyperechoic Renal Masses: Differentiation of Angiomyolipomas from Renal Cell Carcinomas using Tumor Size and Ultrasound Radiomics

A retrospective single-center study was performed to assess the performance of ultrasound image-based texture analysis in differentiating angiomyolipoma (AML) from renal cell carcinoma (RCC) on incidental hyperechoic renal lesions. Ultrasound reports of patients from 2012 to 2017 were queried, and those with a hyperechoic renal mass <5 cm in diameter with further imaging characterization and/or pathological correlation were included. Quantitative texture analysis was performed using a model including 18 texture features. Univariate logistic regression was used to identify texture variables differing significantly between AML and RCC, and the performance of the model was measured using the area under the receiver operating characteristic (ROC) curve. One hundred thirty hyperechoic renal masses in 127 patients characterized as RCCs (25 [19%]) and AMLs (105 [81%]) were included. Size (odds ratio [OR] = 0.12, 95% confidence interval [CI]: 0.04-0.43, p < 0.001) and 4 of 18 texture features, including entropy (OR = 0.09, 95% CI: 0.01-0.81, p = 0.03), gray-level non-uniformity (OR = 0.12, 95% CI: 0.02-0.72, p = 0.02), long-run emphasis (OR = 0.49, 95% CI: 0.27-0.91, p = 0.02) and run-length non-uniformity (OR = 2.18, 95% CI: 1.14-4.16, p = 0.02) were able to differentiate AMLs from RCCs. The area under the ROC curve for the performance of the model, including texture features and size, was 0.945 (p < 0.001). Ultrasound image-based textural analysis enables differentiation of hyperechoic RCCs from AMLs with high accuracy, which improves further when combined with tumor size.