Ultrasound-Based Deep Learning Radiomics Nomogram for the Assessment of Lymphovascular Invasion in Invasive Breast Cancer: A Multicenter Study.

RATIONALE AND OBJECTIVES: The aim of this study was to develop a deep learning radiomics nomogram (DLRN) based on B-mode ultrasound (BMUS) and color doppler flow imaging (CDFI) images for preoperative assessment of lymphovascular invasion (LVI) status in invasive breast cancer (IBC). MATERIALS AND METHODS: In this multicenter, retrospective study, 832 pathologically confirmed IBC patients were recruited from eight hospitals. The samples were divided into training, internal test, and external test sets. Deep learning and handcrafted radiomics features reflecting tumor phenotypes on BMUS and CDFI images were extracted. The BMUS score and CDFI score were calculated after radiomics feature selection. Subsequently, a DLRN was developed based on the scores and independent clinic-ultrasonic risk variables. The performance of the DLRN was evaluated for calibration, discrimination, and clinical usefulness. RESULTS: The DLRN predicted the LVI with accuracy, achieving an area under the receiver operating characteristic curve of 0.93 (95% CI 0.90-0.95), 0.91 (95% CI 0.87-0.95), and 0.91 (95% CI 0.86-0.94) in the training, internal test, and external test sets, respectively, with good calibration. The DLRN demonstrated superior performance compared to the clinical model and single scores across all three sets (p < 0.05). Decision curve analysis and clinical impact curve confirmed the clinical utility of the model. Furthermore, significant enhancements in net reclassification improvement (NRI) and integrated discrimination improvement (IDI) indicated that the two scores could serve as highly valuable biomarkers for assessing LVI. CONCLUSION: The DLRN exhibited strong predictive value for LVI in IBC, providing valuable information for individualized treatment decisions.

Predicting Ki-67 expression in hepatocellular carcinoma: nomogram based on clinical factors and contrast-enhanced ultrasound radiomics signatures.

PURPOSE: To develop a contrast-enhanced ultrasound (CEUS) clinic-radiomics nomogram for individualized assessment of Ki-67 expression in hepatocellular carcinoma (HCC). METHODS: A retrospective cohort comprising 310 HCC individuals who underwent preoperative CEUS (using SonoVue) at three different centers was partitioned into a training set, a validation set, and an external test set. Radiomics signatures indicating the phenotypes of the Ki-67 were extracted from multiphase CEUS images. The radiomics score (Rad-score) was calculated accordingly after feature selection and the radiomics model was constructed. A clinic-radiomics nomogram was established utilizing multiphase CEUS Rad-score and clinical risk factors. A clinical model only incorporated clinical factors was also developed for comparison. Regarding clinical utility, calibration, and discrimination, the predictive efficiency of the clinic-radiomics nomogram was evaluated. RESULTS: Seven radiomics signatures from multiphase CEUS images were selected to calculate the Rad-score. The clinic-radiomics nomogram, comprising the Rad-score and clinical risk factors, indicated a good calibration and demonstrated a better discriminatory capacity compared to the clinical model (AUCs: 0.870 vs 0.797, 0.872 vs 0.755, 0.856 vs 0.749 in the training, validation, and external test set, respectively) and the radiomics model (AUCs: 0.870 vs 0.752, 0.872 vs 0.733, 0.856 vs 0.729 in the training, validation, and external test set, respectively). Furthermore, both the clinical impact curve and the decision curve analysis displayed good clinical application of the nomogram. CONCLUSION: The clinic-radiomics nomogram constructed from multiphase CEUS images and clinical risk parameters can distinguish Ki-67 expression in HCC patients and offer useful insights to guide subsequent personalized treatment.

An overview of ultrasound-derived radiomics and deep learning in liver.

Over the past few years, developments in artificial intelligence (AI), especially in radiomics and deep learning, have enabled the extraction of pathophysiology-related information from varied medical imaging and are progressively transforming medical practice. AI applications are extending into domains previously thought to be accessible only to human experts. Recent research has demonstrated that ultrasound-derived radiomics and deep learning represent an enticing opportunity to benefit preoperative evaluation and prognostic monitoring of diffuse and focal liver disease. This review summarizes the application of radiomics and deep learning in ultrasound liver imaging, including identifying focal liver lesions and staging of liver fibrosis, as well as the evaluation of pathobiological properties of malignant tumors and the assessment of recurrence and prognosis. Besides, we identify important hurdles that must be overcome while also discussing the challenges and opportunities of radiomics and deep learning in clinical applications.

Comparison of microwave alone and combined with ethanol ablation for different types of benign mixed thyroid nodules.

OBJECTIVE: To evaluate the efficacy and safety of microwave ablation (MWA) plus ethanol ablation (EA) for different types of benign mixed thyroid nodules. METHODS: A total of 81 patients with 81 benign mixed thyroid nodules were enrolled into the study; 39 were divided to the MWA group and 42 to the combined group (MWA combined with EA). Nodule ablation rate, volume reduction rate (VRR) and surgical complications of all patients were analyzed before and after treatment. RESULTS: The mean ablation rate were 86.49 ± 6.68% and 90.09 ± 5.79% in the microwave and combined groups respectively, and the ablation rate of nodule decreased as the nodule volume increased. For nodules ≥15 ml in volume, the mean ablation rate of the combined group was higher than that of the microwave group (all P < 0.05). The mean VRR at 12 months postoperatively was 89.58 ± 4.32% in the microwave group and 92.92 ± 3.49% in the combined group, showing statistical significantly different between both arms (P = 0.001). The combined group decreased in volume more significantly than the microwave group for nodules with 20-50% or 50-80% cystic proportions or >15 ml in volume (all P < 0.05). The complication rate was 23.08% and 2.38% respectively. CONCLUSION: MWA combined with EA is more effective than MWA for treating mixed thyroid nodules. MWA combined with EA may be the first approach for nodules with >20% cystic proportions or volume >15 ml.

Evaluation of Tissue Stiffness Around Lesions by Sound Touch Shear Wave Elastography in Breast Malignancy Diagnosis.

The aim of the study described here was to assess the evaluation of tissue stiffness around lesions by sound touch shear wave elastography (STE) in breast malignancy diagnosis. This was an institutional ethics committee-approved, single-center study. A total of 90 women with breast masses examined with conventional ultrasound and STE were eligible for enrollment from December 2020 to July 2021. The maximum and mean elastic values of masses, Emax and Emean, were determined. Shell function was used to measure the maximum and mean elastic values of tissues around masses in annular shells 0.5, 1.0, 1.5 and 2.0 mm wide, recorded as corresponding Emax-shell and Emean-shell. All parameters were analyzed and compared with histopathologic results. Receiver operating characteristic curves were constructed to assess diagnostic performance. Logistic regression analysis was conducted to determine the best diagnostic model. Collagen fiber content of tissues around breast lesions was evaluated using Masson staining and ImageJ software. Ninety women with breast masses were included in this study; 50 had benign (mean diameter 15.84 ± 4.39 mm) and 40 had malignant (mean diameter 17.40 ± 5.42 mm) masses. The diagnostic value of Emax-shell-2.0 was the highest (area under the curve = 0.930) with a sensitivity of 87.5% and specificity of 88%. According to stepwise logistic regression analysis, Emax-shell-2.0 and age were independent predictors of malignancy. Emax-shell-2.0 was also found to be highly correlated with the collagen fiber content of tissue in the malignant group (r = 0.877). Tissue stiffness around lesions measured by STE is a useful metric in identifying malignant breast masses by reflecting collagen fiber content, and Emax-shell-2.0 performs best.

Preoperative prediction of axillary lymph node metastasis in patients with breast cancer based on radiomics of gray-scale ultrasonography.

BACKGROUND: To investigate the performance of a radiomics model based on gray-scale ultrasonography (US) for the preoperative non-invasive prediction of ipsilateral axillary lymph node (ALN) metastasis in patients with breast cancer (BC). METHODS: A total of 192 pathologically confirmed BC patients were included in this study. The training set was comprised of 132 patients from hospital 1 and the test set was comprised of 60 patients from hospital 2. All patients underwent US before percutaneous core biopsy and the results of ALN status reported by a radiologist with 12 years of experience were recorded. Radiomic features were extracted from the gray-scale US images. Max-relevance and min-redundancy (MRMR) and least absolute shrinkage and selection operator (LASSO) were used for data dimension reduction and feature selection. A radiomics model was constructed using LASSO and was validated using the leave group out cross-validation (LGOCV) method. The performance of the model was validated with receiver operating characteristic (ROC), calibration curve, and decision curve analysis. RESULTS: A total of 860 features were extracted from the gray-scale US images of each breast lesion, and 9 radiomic features were selected for model construction. The area under the curve (AUC), sensitivity, and specificity of the model for predicting ALN metastasis were 0.85, 78.9%, and 77.3% in the training set and 0.65, 68.0%, and 79.4% in the test set, respectively. The prediction performance of the model was significantly higher than that of the radiologist (AUC: 0.85 vs. 0.59, P<0.01) in the training set and was slightly higher than that of the radiologist (AUC: 0.65 vs. 0.63, P>0.05) in the test set. CONCLUSIONS: The non-invasive radiomics model has the ability to predict ALN metastasis for patients with breast cancer and may outperform US-reported ALN status performed by the radiologist.

Predict esophageal varices via routine trans-abdominal ultrasound: A design of classification analysis model.

BACKGROUND AND AIMS: Upper gastrointestinal endoscopy remains the gold standard for diagnosis of esophageal varices. Trans-abdominal ultrasound, as a noninvasive routine examination for the follow-up of cirrhosis patient, is safe, cheap, easy to perform, and plays an important role. In this study, we attempt to design a practical classification analysis model to predict esophageal varices via ultrasound. METHODS: Compared with endoscopy, the ultrasound qualitative signs (lower esophageal Doppler signals, left gastric vein hepatofugal flow, and paraumbilical vein recanalization) and quantitative parameters (spleen diameter, spleen vein diameter, portal vein diameter, and portal vein velocity) have been evaluated in 286 cirrhosis patients. RESULTS: The classification analysis model is designed as that: the patients are defined with esophageal varices high risk, who with any ultrasound qualitative signs or who with spleen diameter greater than 162 mm without qualitative parameters. The sensitivity for detecting esophageal varices is 97.5% and the specificity is 82.6%, while the positive predictive value is 96.7%, negative predictive value is 83.4%, and the omission diagnostic rate is 2.5%. CONCLUSIONS: This classification analysis model design includes ultrasound qualitative signs and spleen diameter, which can be detected easily via routine ultrasound without other auxiliary. The classification analysis model is useful in detecting esophageal varices, which may be a supplement for predicting of esophageal varices, and reducing the frequency of endoscopy in the follow-up of cirrhosis patients.

Esophageal varix predictive performance of lower esophageal Doppler signals during the swallowing process.

The objective of this study was to assess whether the swallowing action can improve the display of lower esophageal Doppler signals (LEDS) during transabdominal ultrasound (TUS). Eighty-four patients with cirrhosis underwent both TUS and endoscopic examination for esophageal varices (EVs). LEDS were assessed under the esophageal resting state and during the swallowing process. Univariate analysis indicated that spleen diameter, spleen vein diameter, portal vein diameter, LEDS and left gastric vein hepatofugal flow were significantly associated with the presence of EVs. No LEDS were detected in patients without EVs at rest or during swallowing. Of the 69 patients with EVs, LEDS could be detected in 21 cases (30.4%) in the esophageal resting state and in 58 cases (84.1%) during the swallowing process. Compared with the esophageal resting state, the swallowing action can significantly improve display of LEDS during TUS (p = 0.000), which may be beneficial for TUS detection of EVs.

Noninvasive analysis of portal pressure by contrast-enhanced sonography in patients with cirrhosis.

OBJECTIVES: Free portal pressure measurement is a reliable method for assessment of portal pressure in patients with cirrhosis. Intrahepatic circulatory time analysis of a sonographic contrast agent can assess liver fibrosis and its severity. The purposes of this pilot study were to assess the correlation between the intrahepatic circulatory time and free portal pressure and to assess whether intrahepatic circulatory time analysis can be used to predict portal venous pressure severity. METHODS: The intrahepatic circulatory time and free portal pressure were measured in 31 patients with hepatitis B virus-related liver disease. Pearson correlation analysis was used to assess the correlation between the intrahepatic circulatory time and free portal pressure. RESULTS: The hepatic vein-hepatic artery interval times were significantly shorter in the portal hypertension group than the non-portal hypertension group (mean ± SD, 8.26 ± 1.94 and 13.83 ± 1.17 seconds, respectively; P < .001). The portal vein-hepatic artery interval times were significantly longer in the portal hypertension group than the nonportal hypertension group (13.13 ± 2.25 and 7.25 ± 1.81 seconds; P < .001). Considering the whole patient population, there were statistically significant correlations between free portal pressure and the hepatic vein-hepatic artery interval time (r = -0.900; P < .001) and portal vein-hepatic artery interval time (r = 0.808; P < .001). In patients with portal hypertension, there was a statistically significant correlation between free portal pressure and the hepatic vein-hepatic artery interval time (r = -0.804; P = .009) and a weak correlation between free portal pressure and the portal vein-hepatic artery interval time (r = 0.506; P = .036). CONCLUSIONS: Intrahepatic circulatory time measurement is correlated with free portal pressure and has the potential capability to evaluate portal pressure noninvasively in patients with hepatitis B virus-related liver disease.