Preliminary Clinical Experience with a Novel Optical–Ultrasound Imaging Device on Various Skin Lesions

A compact handheld skin ultrasound imaging device has been developed that uses co-registered optical and ultrasound imaging to provide diagnostic information about the full skin depth. The aim of the current work is to present the preliminary clinical results of this device. Using additional photographic, dermoscopic and ultrasonic images as reference, the images from the device were assessed in terms of the detectability of the main skin layer boundaries and characteristic image features. Combined optical-ultrasonic recordings of various types of skin lesions (melanoma, basal cell carcinoma, seborrheic keratosis, dermatofibroma, naevus, dermatitis and psoriasis) were taken with the device (N = 53) and compared with images captured with a reference portable skin ultrasound imager. The investigator and two additional independent experts performed the evaluation. The detectability of skin structures was over 90% for the epidermis, the dermis and the lesions. The morphological and echogenicity information observed for the different skin lesions were found consistent with those of the reference ultrasound device and relevant ultrasound images in the literature. The presented device was able to obtain simultaneous in-vivo optical and ultrasound images of various skin lesions. This has the potential for further investigations, including the preoperative planning of skin cancer treatment.

Evaluation of contrast enhancement ultrasound images of Sonazoid microbubbles in tissue-mimicking phantom obtained by optimal Golay pulse compression

In contrast enhancement ultrasound (CEUS), the vasculature image can be formed from nonlinear echoes arising from microbubbles in a blood flow. The use of binary-coded pulse compression is promising for improving the contrast of CEUS images by suppressing background noise. However, the amplitudes of nonlinear echoes can be reduced, and sidelobes by nonlinear echoes can occur depending on the binary code. Optimal Golay codes with slight nonlinear-echo reduction and nonlinear sidelobe have been proposed. In this study, CEUS images obtained by optimal Golay pulse compression are evaluated through experiments using Sonazoid microbubbles flowing in a tissue-mimicking phantom.

Master Frame Extraction of Fetal Cardiac Images Using B Mode Ultrasound Images

Fetal Echocardiography is used for monitoring the fetal heart and for detection of Congenital Heart Disease (CHD). It is well known that fetal cardiac four chamber view has been widely used for preliminary examination for the detection of CHD. The end diastole frame is generally used for the analysis of the fetal cardiac chambers which is manually picked by the clinician during examination/screening. This method is subjected to intra and inter observer errors and also time consuming. The proposed study aims to automate this process by determining the frame, referred to as the Master frame from the cine loop sequences that can be used for the analysis of the fetal heart chambers instead of the clinically chosen diastole frame. The proposed framework determines the correlation between the reference (first) frame with the successive frames to identify one cardiac cycle. Then the Master frame is formed by superimposing all the frames belonging to one cardiac cycle. The master frame is then compared with the clinically chosen diastole frame in terms of fidelity metrics such as Dice coefficient, Hausdorff distance, mean square error and structural similarity index. The average value of the fidelity metrics considering the dataset used for this study 0.73 for Dice, 13.94 for Hausdorff distance, 0.99 for Structural Similarity Index and 0.035 for mean square error confirms the suitability of the proposed master frame extraction thereby avoiding manual intervention by the clinician. .

Automatic Deep Learning-Based Consolidation/Collapse Classification in Lung Ultrasound Images for COVID-19 Induced Pneumonia

<div>Our automated deep learning-based approach identifies consolidation/collapse in LUS images to aid in the diagnosis of late stages of COVID-19 induced pneumonia, where consolidation/collapse is one of the possible associated pathologies. A common challenge in training such models is that annotating each frame of an ultrasound video requires high labelling effort. This effort in practice becomes prohibitive for large ultrasound datasets. To understand the impact of various degrees of labelling precision, we compare labelling strategies to train fully supervised models (frame-based method, higher labelling effort) and inaccurately supervised models (video-based methods, lower labelling effort), both of which yield binary predictions for LUS videos on a frame-by-frame level. We moreover introduce a novel sampled quaternary method which randomly samples only 10% of the LUS video frames and subsequently assigns (ordinal) categorical labels to all frames in the video based on the fraction of positively annotated samples. This method outperformed the inaccurately supervised video-based method of our previous work on pleural effusions. More surprisingly, this method outperformed the supervised frame-based approach with respect to metrics such as precision-recall area under curve (PR-AUC) and F1 score that are suitable for the class imbalance scenario of our dataset despite being a form of inaccurate learning. This may be due to the combination of a significantly smaller data set size compared to our previous work and the higher complexity of consolidation/collapse compared to pleural effusion, two factors which contribute to label noise and overfitting; specifically, we argue that our video-based method is more robust with respect to label noise and mitigates overfitting in a manner similar to label smoothing. Using clinical expert feedback, separate criteria were developed to exclude data from the training and test sets respectively for our ten-fold cross validation results, which resulted in a PR-AUC score of 73% and an accuracy of 89%. While the efficacy of our classifier using the sampled quaternary method must be verified on a larger consolidation/collapse dataset, when considering the complexity of the pathology, our proposed classifier using the sampled quaternary video-based method is clinically comparable with trained experts and improves over the video-based method of our previous work on pleural effusions.</div>

Preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma based on kupffer phase radiomic features of sonazoid contrast-enhanced ultrasound (SCEUS): A prospective study

OBJECTIVES: To establish and evaluate a machine learning radiomics model based on grayscale and Sonazoid contrast enhanced ultrasound images for the preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) patients. METHODS: 100 cases of histopathological confirmed HCC lesions were prospectively included. Regions of interest were segmented on both grayscale and Kupffer phase of Sonazoid contrast enhanced (CEUS) images. Radiomic features were extracted from tumor region and region containing 5 mm of peritumoral liver tissues. Maximum relevance minimum redundancy (MRMR) and Least Absolute Shrinkage and Selection Operator (LASSO) were used for feature selection and Support Vector Machine (SVM) classifier was trained for radiomic signature calculation. Radiomic signatures were incorporated with clinical variables using univariate-multivariate logistic regression for the final prediction of MVI. Receiver operating characteristic curves, calibration curves and decision curve analysis were used to evaluate model’s predictive performance of MVI. RESULTS: Age were the only clinical variable significantly associated with MVI. Radiomic signature derived from Kupffer phase images of peritumoral liver tissues (kupfferPT) displayed a significantly better performance with an area under the receiver operating characteristic curve (AUROC) of 0.800 (95% confidence interval: 0.667, 0.834), the final prediction model using Age and kupfferPT achieved an AUROC of 0.804 (95% CI: 0.723, 0.878), accuracy of 75.0%, sensitivity of 87.5% and specificity of 69.1%. CONCLUSIONS: Radiomic model based on Kupffer phase ultrasound images of tissue adjacent to HCC lesions showed an observable better predictive value compared to grayscale images and has potential value to facilitate preoperative identification of HCC patients at higher risk of MVI.

Automatic Deep Learning-Based Consolidation/Collapse Classification in Lung Ultrasound Images for COVID-19 Induced Pneumonia

<div>Our automated deep learning-based approach identifies consolidation/collapse in LUS images to aid in the diagnosis of late stages of COVID-19 induced pneumonia, where consolidation/collapse is one of the possible associated pathologies. A common challenge in training such models is that annotating each frame of an ultrasound video requires high labelling effort. This effort in practice becomes prohibitive for large ultrasound datasets. To understand the impact of various degrees of labelling precision, we compare labelling strategies to train fully supervised models (frame-based method, higher labelling effort) and inaccurately supervised models (video-based methods, lower labelling effort), both of which yield binary predictions for LUS videos on a frame-by-frame level. We moreover introduce a novel sampled quaternary method which randomly samples only 10% of the LUS video frames and subsequently assigns (ordinal) categorical labels to all frames in the video based on the fraction of positively annotated samples. This method outperformed the inaccurately supervised video-based method of our previous work on pleural effusions. More surprisingly, this method outperformed the supervised frame-based approach with respect to metrics such as precision-recall area under curve (PR-AUC) and F1 score that are suitable for the class imbalance scenario of our dataset despite being a form of inaccurate learning. This may be due to the combination of a significantly smaller data set size compared to our previous work and the higher complexity of consolidation/collapse compared to pleural effusion, two factors which contribute to label noise and overfitting; specifically, we argue that our video-based method is more robust with respect to label noise and mitigates overfitting in a manner similar to label smoothing. Using clinical expert feedback, separate criteria were developed to exclude data from the training and test sets respectively for our ten-fold cross validation results, which resulted in a PR-AUC score of 73% and an accuracy of 89%. While the efficacy of our classifier using the sampled quaternary method must be verified on a larger consolidation/collapse dataset, when considering the complexity of the pathology, our proposed classifier using the sampled quaternary video-based method is clinically comparable with trained experts and improves over the video-based method of our previous work on pleural effusions.</div>