Correlation between oxygenation status of extra-synovial tissue of wrist and disease activity in rheumatoid arthritis: a photoacoustic imaging study.

OBJECTIVES: Rheumatoid arthritis (RA) is characterized by hypoxia in the synovial tissue. While photoacoustic imaging (PA) offers a method to evaluate tissue oxygenation in RA patients, studies exploring the link between extra-synovial tissue of wrist oxygenation and disease activity remain scarce. We aimed to assess synovial oxygenation in RA patients using a multimodal photoacoustic-ultrasound (PA/US) imaging system and establish its correlation with disease activity. METHODS: A retrospective study was conducted on 111 patients with RA and 72 healthy controls from 2022 to 2023. Dual-wavelength PA imaging quantified oxygen saturation (So2) levels in the synovial membrane and peri-wrist region. Oxygenation states were categorised as hyperoxia, intermediate oxygenation, and hypoxia based on So2 values. The association between oxygenation levels and the clinical disease activity index was evaluated using a one-way analysis of variance, complemented by the Kruskal-Wallis test with Bonferroni adjustment. RESULTS: Of the patients with RA, 39 exhibited hyperoxia, 24 had intermediate oxygenation, and 48 had hypoxia in the wrist extra-synovial tissue. All of the control participants exhibited the hyperoxia status. Oxygenation levels in patients with RA correlated with clinical metrics. Patients with intermediate oxygenation had a lower disease activity index compared with those with hypoxia and hyperoxia. CONCLUSION: A significant correlation exists between wrist extra-synovial tissue oxygenation and disease activity in patients with RA.

Characteristics and Correlations of Wall Shear Stress and Flow Turbulence in the Carotid Bifurcation Evaluated Using an Ultrasound Vector Flow Imaging.

INTRODUCTION: The aim of the study was to evaluate characteristics and provide the normal values of wall shear stress (WSS) and flow turbulence (Tur), and the relationship between them in the carotid bifurcation based on an ultrasound vector flow imaging (V Flow) in healthy adults. METHODS: Max and mean WSS and Tur values at three segments (initial segments of internal and external carotid arteries [IICA and IECA]; distal segment of common carotid artery [DCCA]), both in anterior and posterior walls, were successfully obtained in 56 healthy adults, using ultrasound V Flow function. Relationship between mean WSS and Tur was further explored. RESULTS: The mean WSS value was 0.71 Pa, 0.86 Pa, and 0.96 Pa at IICA, IECA, and DCCA, respectively (IICA < IECA < DCCA, p < 0.05). The mean Tur value was 13.85%, 5.46%, and 4.17% at IICA, IECA, and DCCA, respectively (IICA > IECA > DCCA, p < 0.05). A cutoff value (WSS = 0.4 Pa) was selected and Tur values were significantly higher in group with WSS cutoff value <0.4 Pa than group with WSS cutoff value ≥0.4 Pa (p < 0.01). CONCLUSION: WSS and Tur are moderately negatively correlated, which can be used in the quantitative evaluation of carotid bifurcation and could be a potential dual-parameter tool in the clinical research for early detection of carotid atherosclerosis.

Deep learning to assist composition classification and thyroid solid nodule diagnosis: a multicenter diagnostic study.

OBJECTIVES: This study aimed to propose a deep learning (DL)-based framework for identifying the composition of thyroid nodules and assessing their malignancy risk. METHODS: We conducted a retrospective multicenter study using ultrasound images from four hospitals. Convolutional neural network (CNN) models were constructed to classify ultrasound images of thyroid nodules into solid and non-solid, as well as benign and malignant. A total of 11,201 images of 6784 nodules were used for training, validation, and testing. The area under the receiver-operating characteristic curve (AUC) was employed as the primary evaluation index. RESULTS: The models had AUCs higher than 0.91 in the benign and malignant grading of solid thyroid nodules, with the Inception-ResNet AUC being the highest at 0.94. In the test set, the best algorithm for identifying benign and malignant thyroid nodules had a sensitivity of 0.88, and a specificity of 0.86. In the human vs. DL test set, the best algorithm had a sensitivity of 0.93, and a specificity of 0.86. The Inception-ResNet model performed better than the senior physicians (p < 0.001). The sensitivity and specificity of the optimal model based on the external test set were 0.90 and 0.75, respectively. CONCLUSIONS: This research demonstrates that CNNs can assist thyroid nodule diagnosis and reduce the rate of unnecessary fine-needle aspiration (FNA). CLINICAL RELEVANCE STATEMENT: High-resolution ultrasound has led to increased detection of thyroid nodules. This results in unnecessary fine-needle aspiration and anxiety for patients whose nodules are benign. Deep learning can solve these problems to some extent. KEY POINTS: • Thyroid solid nodules have a high probability of malignancy. • Our models can improve the differentiation between benign and malignant solid thyroid nodules. • The differential performance of one model was superior to that of senior radiologists. Applying this could reduce the rate of unnecessary fine-needle aspiration of solid thyroid nodules.

Comparative study of ultrasound attenuation analysis and controlled attenuation parameter in the diagnosis and grading of liver steatosis in non-alcoholic fatty liver disease patients.

PURPOSE: To assess the diagnostic performance of Ultrasound Attenuation Analysis (USAT) in the diagnosis and grading of hepatic steatosis in patients with non-alcoholic fatty liver disease (NAFLD) using Controlled Attenuation Parameters (CAP) as a reference. MATERIALS AND METHODS: From February 13, 2023, to September 26, 2023, participants underwent CAP and USAT examinations on the same day. We used manufacturer-recommended CAP thresholds to categorize the stages of hepatic steatosis: stage 1 (mild) - 240 dB/m, stage 2 (moderate) - 265 dB/m, stage 3 (severe) - 295 dB/m. Receiver Operating Characteristic curves were employed to evaluate the diagnostic accuracy of USAT and determine the thresholds for different levels of hepatic steatosis. RESULTS: Using CAP as the reference, we observed that the average USAT value increased with the severity of hepatic steatosis, and the differences in USAT values among the different hepatic steatosis groups were statistically significant (p < 0.05). There was a strong positive correlation between USAT and CAP (r = 0.674, p < 0.0001). When using CAP as the reference, the optimal cut-off values for diagnosing and predicting different levels of hepatic steatosis with USAT were as follows: the cut-off value for excluding the presence of hepatic steatosis was 0.54 dB/cm/MHz (AUC 0.96); for mild hepatic steatosis, it was 0.59 dB/cm/MHz (AUC 0.86); for moderate hepatic steatosis, it was 0.73 dB/cm/MHz (AUC 0.81); and for severe hepatic steatosis, it was 0.87 dB/cm/MHz (AUC 0.87). CONCLUSION: USAT exhibits strong diagnostic performance for hepatic steatosis and shows a high correlation with CAP values.

The association between mammographic density and breast cancer risk in Chinese women: a systematic review and meta-analysis.

PURPOSE: Breast density has consistently been shown to be an independent risk factor for breast cancer in Western populations; however, few studies have evaluated this topic in Chinese women and there is not yet a unified view. This study investigated the association between mammographic density (MD) and breast cancer risk in Chinese women. METHODS: The PubMed, Cochrane Library, Embase, and Wanfang databases were systematically searched in June 2023 to include all studies on the association between MD and breast cancer risk in Chinese women. A total of 13,977 breast cancer cases from 14 studies were chosen, including 10 case-control/cross-sectional studies, and 4 case-only studies. For case-control/cross-sectional studies, the odds ratios (ORs) of MD were combined using random effects models, and for case-only studies, relative odds ratios (RORs) were combinations of premenopausal versus postmenopausal breast cancer cases. RESULTS: Women with BI-RADS density category II-IV in case-control/cross-sectional studies had a 0.93-fold (95% confidence interval [CI] 0.55, 1.57), 1.08-fold (95% CI 0.40, 2.94), and 1.24-fold (95% CI 0.42, 3.69) higher risk compared to women with the lowest density category. Combined RORs for premenopausal versus postmenopausal women in case-only studies were 3.84 (95% CI 2.92, 5.05), 22.65 (95% CI 7.21, 71.13), and 42.06 (95% CI 4.22, 419.52), respectively, for BI-RADS density category II-IV versus I. CONCLUSIONS: For Chinese women, breast cancer risk is weakly associated with MD; however, breast cancer risk is more strongly correlated with mammographic density in premenopausal women than postmenopausal women. Further research on the factors influencing MD in premenopausal women may provide meaningful insights into breast cancer prevention in China.

Development and validation of nomograms using photoacoustic imaging and 2D ultrasound to predict breast nodule benignity and malignancy.

BACKGROUND: The application of photoacoustic imaging (PAI), utilizing laser-induced ultrasound, shows potential in assessing blood oxygenation in breast nodules. However, its effectiveness in distinguishing between malignant and benign nodules remains insufficiently explored. PURPOSE: This study aims to develop nomogram models for predicting the benign or malignant nature of breast nodules using PAI. METHOD: A prospective cohort study enrolled 369 breast nodules, subjecting them to PAI and ultrasound examination. The training and testing cohorts were randomly divided into two cohorts in a ratio of 3:1. Based on the source of the variables, three models were developed, Model 1: photoacoustic-BIRADS+BMI + blood oxygenation, Model 2: BIRADS+Shape+Intranodal blood (Doppler) + BMI, Model 3: photoacoustic-BIRADS+BIRADS+ Shape+Intranodal blood (Doppler) + BMI + blood oxygenation. Risk factors were identified through logistic regression, resulting in the creation of three predictive models. These models were evaluated using calibration curves, subject receiver operating characteristic (ROC), and decision curve analysis. RESULTS: The area under the ROC curve for the training cohort was 0.91 (95% confidence interval, 95% CI: 0.88-0.95), 0.92 (95% CI: 0.89-0.95), and 0.97 (95% CI: 0.96-0.99) for Models 1-3, and the ROC curve for the testing cohort was 0.95 (95% CI: 0.91-0.98), 0.89 (95% CI: 0.83-0.96), and 0.97 (95% CI: 0.95-0.99) for Models 1-3. CONCLUSIONS: The calibration curves demonstrate that the model's predictions agree with the actual values. Decision curve analysis suggests a good clinical application.

An artificial intelligence model based on transrectal ultrasound images of biopsy needle tract tissues to differentiate prostate cancer.

PURPOSE: We aimed to develop an artificial intelligence (AI) model based on transrectal ultrasonography (TRUS) images of biopsy needle tract (BNT) tissues for predicting prostate cancer (PCa) and to compare the PCa diagnostic performance of the radiologist model and clinical model. METHODS: A total of 1696 2D prostate TRUS images were involved from 142 patients between July 2021 and May 2022. The ResNet50 network model was utilized to train classification models with different input methods: original image (Whole model), BNT (Needle model), and combined image [Feature Pyramid Networks (FPN) model]. The training set, validation set, and test set were randomly assigned, then randomized 5-fold cross-validation between the training set and validation set was performed. The diagnostic effectiveness of AI models and image combination was accessed by an independent testing set. Then, the optimal AI model and image combination were selected to compare the diagnostic efficacy with that of senior radiologists and the clinical model. RESULTS: In the test set, the area under the curve, specificity, and sensitivity of the FPN model were 0.934, 0.966, and 0.829, respectively; the diagnostic efficacy was improved compared with the Whole and Needle models, with statistically significant differences (P < 0.05), and was better than that of senior radiologists (area under the curve: 0.667). The FPN model detected more PCa compared with senior physicians (82.9% vs. 55.8%), with a 61.3% decrease in the false-positive rate and a 23.2% increase in overall accuracy (0.887 vs. 0.655). CONCLUSION: The proposed FPN model can offer a new method for prostate tissue classification, improve the diagnostic performance, and may be a helpful tool to guide prostate biopsy.

A deep learning-based calculation system for plaque stenosis severity on common carotid artery of ultrasound images.

OBJECTIVES: Assessment of plaque stenosis severity allows better management of carotid source of stroke. Our objective is to create a deep learning (DL) model to segment carotid intima-media thickness and plaque and further automatically calculate plaque stenosis severity on common carotid artery (CCA) transverse section ultrasound images. METHODS: Three hundred and ninety images from 376 individuals were used to train (235/390, 60%), validate (39/390, 10%), and test (116/390, 30%) on a newly proposed CANet model. We also evaluated the model on an external test set of 115 individuals with 122 images acquired from another hospital. Comparative studies were conducted between our CANet model with four state-of-the-art DL models and two experienced sonographers to re-evaluate the present model's performance. RESULTS: On the internal test set, our CANet model outperformed the four comparative models with Dice values of 95.22% versus 90.15%, 87.48%, 90.22%, and 91.56% on lumen-intima (LI) borders and 96.27% versus 91.40%, 88.94%, 91.19%, and 92.88% on media-adventitia (MA) borders. On the external test set, our model still produced excellent results with a Dice value of 92.41%. Good consistency of stenosis severity calculation was observed between CANet model and experienced sonographers, with Intraclass Correlation Coefficient (ICC) of 0.927 and 0.702, Pearson's Correlation Coefficient of 0.928 and 0.704 on internal and external test set, respectively. CONCLUSIONS: Our CANet model achieved excellent performance in the segmentation of carotid IMT and plaques as well as automated calculation of stenosis severity.

A validation of an entropy-based artificial intelligence for ultrasound data in breast tumors.

BACKGROUND: The application of artificial intelligence (AI) in the ultrasound (US) diagnosis of breast cancer (BCa) is increasingly prevalent. However, the impact of US-probe frequencies on the diagnostic efficacy of AI models has not been clearly established. OBJECTIVES: To explore the impact of using US-video of variable frequencies on the diagnostic efficacy of AI in breast US screening. METHODS: This study utilized different frequency US-probes (L14: frequency range: 3.0-14.0 MHz, central frequency 9 MHz, L9: frequency range: 2.5-9.0 MHz, central frequency 6.5 MHz and L13: frequency range: 3.6-13.5 MHz, central frequency 8 MHz, L7: frequency range: 3-7 MHz, central frequency 4.0 MHz, linear arrays) to collect breast-video and applied an entropy-based deep learning approach for evaluation. We analyzed the average two-dimensional image entropy (2-DIE) of these videos and the performance of AI models in processing videos from these different frequencies to assess how probe frequency affects AI diagnostic performance. RESULTS: The study found that in testing set 1, L9 was higher than L14 in average 2-DIE; in testing set 2, L13 was higher in average 2-DIE than L7. The diagnostic efficacy of US-data, utilized in AI model analysis, varied across different frequencies (AUC: L9 > L14: 0.849 vs. 0.784; L13 > L7: 0.920 vs. 0.887). CONCLUSION: This study indicate that US-data acquired using probes with varying frequencies exhibit diverse average 2-DIE values, and datasets characterized by higher average 2-DIE demonstrate enhanced diagnostic outcomes in AI-driven BCa diagnosis. Unlike other studies, our research emphasizes the importance of US-probe frequency selection on AI model diagnostic performance, rather than focusing solely on the AI algorithms themselves. These insights offer a new perspective for early BCa screening and diagnosis and are of significant for future choices of US equipment and optimization of AI algorithms.

The research on artificial intelligence-assisted breast diagnosis often relies on static images or dynamic videos obtained from ultrasound probes with different frequencies. However, the effect of frequency-induced image variations on the diagnostic performance of artificial intelligence models remains unclear. In this study, we aimed to explore the impact of using ultrasound images with variable frequencies on AI's diagnostic efficacy in breast ultrasound screening. Our approach involved employing a video and entropy-based feature breast network to compare the diagnostic efficiency and average two-dimensional image entropy of the L14 (frequency range: 3.0-14.0 MHz, central frequency 9 MHz), L9 (frequency range: 2.5-9.0 MHz, central frequency 6.5 MHz) linear array probe and L13 (frequency range: 3.6-13.5 MHz, central frequency 8 MHz), and L7 (frequency range: 3-7 MHz, central frequency 4.0 MHz) linear array probes. The results revealed that the diagnostic efficiency of AI models differed based on the frequency of the ultrasound probe. It is noteworthy that ultrasound images acquired with different frequency probes exhibit different average two-dimensional image entropy, while higher average two-dimensional image entropy positively affect the diagnostic performance of the AI model. We concluded that a dataset with higher average two-dimensional image entropy is associated with superior diagnostic efficacy for AI-based breast diagnosis. These findings contribute to a better understanding of how ultrasound image variations impact AI-assisted breast diagnosis, potentially leading to improved breast cancer screening outcomes.

Blood Flow Turbulence Quantification of Carotid Artery With a High-Frame Rate Vector Flow Imaging.

OBJECTIVES: To assess the feasibility and performance of Turbulence (Tur) index as a quantitative tool for carotid artery flow turbulence; to detect and compare the blood flow patterns of common carotid artery (CCA) and carotid bulb (CB) at different ages and cardiac phases in healthy adults, and thus interpret the evolvement of etiology difference between CCA and CB. METHODS: Carotid flow characteristics of 40 healthy volunteers were evaluated quantitatively by a high-frame rate vector flow imaging. Three types of flow patterns were defined depending on the distributive range of complex flow during systole in CB. Comparison of mean Tur value in CCA and CB at different age groups and cardiac phases was performed. And the correlation between Tur value and the diameter ratio of proximal internal carotid artery to common carotid artery (DRpro-ica/cca) was tested. RESULTS: Mean Tur values in CB were remarkably higher than that in CCA, whether during systole or diastole (P < .001). Meanwhile Tur values in CB during systole were significantly higher than that during diastole (P < .001). Flow complexity of CB showed variations among 40 participants especially in systole, whereas the flow pattern of CCA was relatively consistent. Mean Tur values were positively correlated with DRpro-ica/cca in CB (ρ = 0.69, P < .05). CONCLUSIONS: V Flow imaging provided a reliable method-Tur, for quantitative analysis of carotid blood flow. It had potential to be further applied in distinguishing complex hemodynamic characteristics in high-risk people of carotid diseases for the risk stratification of cardiovascular events.

Artificial intelligence for non-mass breast lesions detection and classification on ultrasound images: a comparative study.

BACKGROUND: This retrospective study aims to validate the effectiveness of artificial intelligence (AI) to detect and classify non-mass breast lesions (NMLs) on ultrasound (US) images. METHODS: A total of 228 patients with NMLs and 596 volunteers without breast lesions on US images were enrolled in the study from January 2020 to December 2022. The pathological results served as the gold standard for NMLs. Two AI models were developed to accurately detect and classify NMLs on US images, including DenseNet121_448 and MobileNet_448. To evaluate and compare the diagnostic performance of AI models, the area under the curve (AUC), accuracy, specificity and sensitivity was employed. RESULTS: A total of 228 NMLs patients confirmed by postoperative pathology with 870 US images and 596 volunteers with 1003 US images were enrolled. In the detection experiment, the MobileNet_448 achieved the good performance in the testing set, with the AUC, accuracy, sensitivity, and specificity were 0.999 (95%CI: 0.997-1.000),96.5%,96.9% and 96.1%, respectively. It was no statistically significant compared to DenseNet121_448. In the classification experiment, the MobileNet_448 model achieved the highest diagnostic performance in the testing set, with the AUC, accuracy, sensitivity, and specificity were 0.837 (95%CI: 0.990-1.000), 70.5%, 80.3% and 74.6%, respectively. CONCLUSIONS: This study suggests that the AI models, particularly MobileNet_448, can effectively detect and classify NMLs in US images. This technique has the potential to improve early diagnostic accuracy for NMLs.

Appendicitis in pregnant women: A systematic review and meta-analysis of the diagnostic performance of ultrasonography.

OBJECTIVES: The accuracy of ultrasound in the detection of appendicitis in pregnant women was examined in a meta-analysis. METHODS: Pregnant women with suspected acute appendicitis were evaluated using ultrasound in a systematic search of PubMed, EMBASE, and Cochrane Library databases from January 1, 2011 to August 10, 2023. The sensitivity and specificity values and diagnostic odds ratios were obtained using the pooled data. RESULTS: A total of 239 patients were studied in four relevant investigations. Ultrasonography has a sensitivity of 56% and a specificity of 88% for the diagnosis of acute appendicitis, with an area under the receiver operating characteristic curve of 0.66%. Ultrasonography had a positive likelihood ratio of 4.65 (95% confidence interval, 1.42-15.23) and a negative likelihood ratio of 0.50 (95% confidence interval, 0.41-0.62). There was no evidence of publication bias (p = 0.93). CONCLUSIONS: Ultrasound has moderate sensitivity for identifying appendicitis in pregnant women and may be utilized as an alternative diagnostic method.

Precise evaluation of blood flow patterns in human carotid bifurcation based on high-frame-rate vector flow imaging.

PURPOSE: To investigate the feasibility of high-frame-rate vector flow imaging (HiFR-VFI) compared to ultrasound color Doppler flow imaging (CDFI) for precisely evaluating flow characteristics in the carotid bifurcation (CB) of presumed healthy adults. METHODS: Forty-three volunteers were assessed for flow characteristics and their extensions using HiFR-VFI and CDFI in CBs. The flow patterns were classified according to the streamlines in HiFR-VFI and quantitatively measured using an innovative turbulence index (Tur-value). Interobserver agreement was also assessed. RESULTS: HiFR-VFI was consistent with CDFI in detecting laminar and nonlaminar flow in 81.4% of the cases; however, in 18.6% of the cases, only HiFR-VFI identified the nonlaminar flow. HiFR-VFI showed a larger extension of complex flow (0.37 ± 0.26 cm2 ) compared to CDFI (0.22 ± 0.21 cm2 ; p < 0.05). The flow patterns were classified into four types: 3 type-I (laminar flow), 35 type-II (rotational flow), 27 type-III (reversed flow), and 5 type-IV (complex flow). The Tur-value of type-IV (50.03 ± 14.97)% is larger than type-III (44.57 ± 8.89)%, type-II (16.30 ± 8.16)%, and type-I (1.48 ± 1.43)% (p < 0.05). Two radiologists demonstrated almost perfect interobserver agreement on recognizing the change of streamlines (κ = 0.81, p < 0.001). The intraclass correlation coefficient of the Tur-value was 0.98. CONCLUSION: HiFR-VFI can reliably characterize complex hemodynamics with quantitative turbulence measurement and may be an auxiliary diagnostic tool for assessing atherosclerotic arterial disease.

Echocardiographic study of myocardial work in patients with type 2 diabetes mellitus.

BACKGROUND: A noninvasive left ventricular (LV) pressure-strain loop (PSL) provides a new method to quantify myocardial work (MW) by combining global longitudinal strain (GLS) and LV pressure, which exerts potential advantages over traditional GLS. We studied the LV PSL and MW in patients with type 2 diabetes mellitus (T2DM). METHODS: This cross-sectional study included 201 subjects (54 healthy controls and 147 T2DM patients) who underwent complete two-dimensional echocardiography (2DE), including 2D speckle-tracking echocardiography (STE), as well as brachial artery pulse pressure measurement. The PSL was used to determine the global myocardial work index (GWI), global constructive work (GCW), global wasted work (GWW), and global work efficiency (GWE) of all study participants. The association between T2DM and LV function was evaluated according to these MW indices. RESULTS: The GLS was significantly lower in the T2DM group than in the control group (P < 0.001), indicating that the LV myocardium had been damaged, although the LV ejection fraction (LVEF) was still normal. The GWI and GWE were decreased (P = 0.022) and the GWW was increased (P < 0.001) in diabetic patients compared with controls, but the GCW was comparable in the two groups (P = 0.160). In all diabetic patients, age, body mass index, systolic blood pressure, smoking history, and LVEF were correlated with GWI, GWW and GWE. CONCLUSIONS: The use of LV PSL is a novel noninvasive technique that could help to depict the relationship between LV myocardial damage and MW in patients with T2DM.

A deep learning classification of metacarpophalangeal joints synovial proliferation in rheumatoid arthritis by ultrasound images.

OBJECTIVE: To evaluate if an automatic classification of rheumatoid arthritis (RA) metacarpophalangeal joint conditions in ultrasound images is feasible by deep learning (DL) method, to provide a more objective, automated, and fast way of RA diagnosis in clinical setting. MATERIALS AND METHODS: DenseNet-based DL model was used and both training and testing are implemented in TensorFlow 1.13.1 with Keras DL libraries. The area under curve (AUC), accuracy, sensitivity, and specificity values with 95% CIs were reported. The statistical analysis was performed by using scikit-learn libraries in Python 3.7. RESULTS: A total of 1337 RA ultrasound images were acquired from 208 patients, the number of images is 313, 657, 178, and 189 in OESS Grade L0, L1, L2, and L3, respectively. In Classification Scenario 1 SP-no versus SP-yes, three experiments with region of interest of size 192 × 448 (Group 1), 96 × 224 (Group 2), and 96 × 224 stacked with pre-segmented annotated mask of SP area (Group 3) as input achieve an AUC of 0.863 (95% CI: 0.809, 0.917), 0.861 (95% CI: 0.805, 0.916), and 0.886 (95% CI: 0.836, 0.936), respectively. In Classification Scenario 2 Healthy versus Diseased, experiments in Group 1, Group 2 and Group 3 achieve an AUC of 0.848 (95% CI: 0.799, 0.896), 0.864 (95% CI: 0.819, 0.909), and 0.916 (95% CI: 0.883, 0.952), respectively. CONCLUSION: We combined DenseNet model with ultrasound images for RA condition assessment. The feasibility of using DL to create an automatic RA condition classification system was also demonstrated. The proposed method can be an alternative to the initial screening of RA patients.

Differential diagnostic value of periampullary mass: A nomogram established by random forest based on clinical characteristics and contrast-enhanced ultrasound.

PURPOSES: To develop a nomogram model for distinguishing benign from malignant ampullary lesions more intuitive and accurate. MATERIALS AND METHODS: A total of 124 patients with periampullary lesions from January 2016 to June 2020 were enrolled in this retrospective study. Their clinical information, ultrasound (US), dual contrast-enhanced ultrasound (DCEUS) and MRI image features were used for research. Twenty features were collected in our study. Random forest was used to select the first five most important indicators to construct the prediction model. RESULTS: Patients' age, common bile duct (CBD) diameter, the shape, vascularity, and boundary of lesion, lesion size with or without enlarged after CEUS, the enhancement patterns of arterial phase, the washout patterns of venous phase, CEUS diagnosis, and MRI diagnosis were statistically significant (p < 0.05). After screening for statistically significant indicators by random forest, the first five most important indicators were age, CBD diameter, the enhancement patterns of arterial phase, the washout patterns of venous phase, lesion size with or without enlarged after CEUS, which were used to construct nomogram. The area under curves (AUC) and 95% confidence intervals (CI) for nomogram, MRI + MRCP + DCEUS, DCEUS, MRI + MRCP were 0.98(0.94-1.00), 0.91(0.84-0.97), 0.89(0.80-0.98), 0.68(0.60-0.77), respectively. The sensitivity and specificity were 100.00% and 84.62% for nomogram, 88.29% and 92.31% for MRI + MRCP+DCEUS, 86.49% and 92.31% for DCEUS, 51.35%, and 100.00% for MRI + MRCP. CONCLUSIONS: We combined clinical indicators, gray-scale ultrasound characteristics, and CEUS characteristics to build the nomogram, which can be intuitively and accurately used for preoperative malignant prediction of ampullary lesion patients, worthy of clinical generalizability and application.

One step further into the blackbox: a pilot study of how to build more confidence around an AI-based decision system of breast nodule assessment in 2D ultrasound.

OBJECTIVES: To investigate how a DL model makes decisions in lesion classification with a newly defined region of evidence (ROE) by incorporating "explainable AI" (xAI) techniques. METHODS: A data set of 785 2D breast ultrasound images acquired from 367 females. The DenseNet-121 was used to classify whether the lesion is benign or malignant. For performance assessment, classification results are evaluated by calculating accuracy, sensitivity, specificity, and receiver operating characteristic for experiments of both coarse and fine regions of interest (ROIs). The area under the curve (AUC) was evaluated, and the true-positive, false-positive, true-negative, and false-negative results with breakdown in high, medium, and low resemblance on test sets were also reported. RESULTS: The two models with coarse and fine ROIs of ultrasound images as input achieve an AUC of 0.899 and 0.869, respectively. The accuracy, sensitivity, and specificity of the model with coarse ROIs are 88.4%, 87.9%, and 89.2%, and with fine ROIs are 86.1%, 87.9%, and 83.8%, respectively. The DL model captures ROE with high resemblance of physicians' consideration as they assess the image. CONCLUSIONS: We have demonstrated the effectiveness of using DenseNet to classify breast lesions with limited quantity of 2D grayscale ultrasound image data. We have also proposed a new ROE-based metric system that can help physicians and patients better understand how AI makes decisions in reading images, which can potentially be integrated as a part of evidence in early screening or triaging of patients undergoing breast ultrasound examinations. KEY POINTS: • The two models with coarse and fine ROIs of ultrasound images as input achieve an AUC of 0.899 and 0.869, respectively. The accuracy, sensitivity, and specificity of the model with coarse ROIs are 88.4%, 87.9%, and 89.2%, and with fine ROIs are 86.1%, 87.9%, and 83.8%, respectively. • The first model with coarse ROIs is slightly better than the second model with fine ROIs according to these evaluation metrics. • The results from coarse ROI and fine ROI are consistent and the peripheral tissue is also an impact factor in breast lesion classification.

Promoter hypermethylation of the CFTR gene as a novel diagnostic and prognostic marker of breast cancer.

Breast cancer is the most commonly diagnosed cancer and the leading cause of cancer-related deaths among women. New biomarkers with definite diagnostic and prognostic efficacy are urgently needed. Here, we showed that the promoter of the cystic fibrosis transmembrane conductance regulator (CFTR) was hypermethylated in breast cancer. The messenger RNA level of CFTR was downregulated in breast cancer. Notably, all 19 breast cancer patients with hypermethylated CFTR were diagnosed with invasive carcinoma. Moreover, CFTR was upregulated in decitabine (10 µM) treated breast cancer cells. Overexpression of CFTR inhibited cell growth whereas knockdown of CFTR promoted cell invasion. In the tissue array analysis, the CFTR protein level decreased significantly in breast cancer and low CFTR protein level correlated with poor survival with a P-value of 0.034. Thus, promoter hypermethylation of the CFTR gene might be a novel diagnostic marker of breast cancer.

Clinical Value of the Elastographic Q-Analysis Score in Assisting Real-Time Elastography-Guided Prostate Biopsy: A Retrospective Study of 125 Patients.

OBJECTIVES: This study aimed to evaluate the clinical value of the elastographic Q-analysis score (EQS) in assisting real-time elastography- and transrectal US-guided prostate biopsy. METHODS: A total of 125 patients with 301 lesions were enrolled in this study; all were confirmed by pathologic results. The patients underwent transrectal US and elastographic examinations before biopsy. Elastographic Q-analysis score analysis software was used for measuring the mean EQS of the elastic images. First, the suspicious regions on elastography underwent biopsy. Then 12-core systematic prostate biopsy was performed. An EQS curve was used to calculate the mean EQS, and a receiver operating characteristic curve was drawn to find the cutoff point for the EQS to predict prostate cancer. RESULTS: Of the 301 lesions in this study, 125 were malignant, and 176 were benign. The mean EQS values of benign and malignant lesions ± SD were 1.47 ± 0.75 and 2.98 ± 1.06, respectively. The difference was statistically significant (P < .05). The area under the receiver operating characteristic curve was 0.87. When the cutoff point was 1.95 for diagnosing malignant and benign lesions, the sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, and negative likelihood ratio were 83.5%, 84.4%, 76.8%, 89.2%, 5.35, and 0.20. CONCLUSIONS: The EQS could be used as a way to predict benign and malignant lesions and thus could serve as guidance for adding targeted biopsy.

Evaluation of the Performance of the Ultrasound (US) Elastographic Q-Analysis Score Combined With the Prostate Imaging Reporting and Data System for Malignancy Risk Stratification in Prostate Nodules Based on Transrectal US-Magnetic Resonance Imaging Fusion Imaging.

OBJECTIVES: This study retrospectively evaluated the prognostic performance of the ultrasound elastographic Q-analysis score (EQS) combined with the Prostate Imaging Reporting and Data System (PI-RADS) for malignancy risk stratification in prostate nodules based on transrectal ultrasound-magnetic resonance imaging fusion imaging. METHODS: Sixty-two patients who were suspected to have PCa between October 2017 and May 2018 in our hospital were retrospectively evaluated. The performance of the EQS and PI-RADS was evaluated by patients' receiver operating characteristic curves in differentiating malignant and benign prostate nodules. The combination of the EQS and PI-RADS methods for prostate imaging was evaluated. RESULTS: Sixty-two prostate nodules in 62 patients were included. All of the patients underwent biopsy; 29 cases were prostate cancer, and the rest were benign prostate lesions. Both the EQS and PI-RADS were significantly higher in malignant nodules than in benign nodules. The sensitivity, specificity, area under the curve, positive likelihood ratio, negative likelihood ratio, positive predictive value, negative predictive value, and Youden index of an EQS cutoff of 2.05 were 86.2%, 81.8%, 85.9%, 4.73, 0.169, 80.6%, 87.1%, and 68%, respectively. The corresponding numbers for a PI-RADS cutoff of 4 were 82.7%, 69.7%, 84.2%, 2.72, 0.25, 70.6%, 82.1%, and 52.4%. The "tandem" method had a higher diagnostic specificity (87.9%), positive likelihood ratio (6.55), and positive predictive value (85.1%). The "parallel" method had a higher diagnostic sensitivity (96.5%), negative likelihood ratio (0.06), and negative predictive value (95.2%). CONCLUSIONS: both the EQS and PI-RADS had good diagnostic performance in differentiating between malignant and benign prostate lesions. The combination of the EQS and PI-RADS improved the diagnostic performance to a certain degree.