Context-aware deep learning enables high-efficacy localization of high concentration microbubbles for super-resolution ultrasound localization microscopy.

Ultrasound localization microscopy (ULM) enables deep tissue microvascular imaging by localizing and tracking intravenously injected microbubbles circulating in the bloodstream. However, conventional localization techniques require spatially isolated microbubbles, resulting in prolonged imaging time to obtain detailed microvascular maps. Here, we introduce LOcalization with Context Awareness (LOCA)-ULM, a deep learning-based microbubble simulation and localization pipeline designed to enhance localization performance in high microbubble concentrations. In silico, LOCA-ULM enhanced microbubble detection accuracy to 97.8% and reduced the missing rate to 23.8%, outperforming conventional and deep learning-based localization methods up to 17.4% in accuracy and 37.6% in missing rate reduction. In in vivo rat brain imaging, LOCA-ULM revealed dense cerebrovascular networks and spatially adjacent microvessels undetected by conventional ULM. We further demonstrate the superior localization performance of LOCA-ULM in functional ULM (fULM) where LOCA-ULM significantly increased the functional imaging sensitivity of fULM to hemodynamic responses invoked by whisker stimulations in the rat brain.

Contrast-Free Super-Resolution Power Doppler (CS-PD) Based on Deep Neural Networks.

Super-resolution ultrasound microvessel imaging based on ultrasound localization microscopy (ULM) is an emerging imaging modality that is capable of resolving micrometer-scaled vessels deep into tissue. In practice, ULM is limited by the need for contrast injection, long data acquisition, and computationally expensive postprocessing times. In this study, we present a contrast-free super-resolution power Doppler (CS-PD) technique that uses deep networks to achieve super-resolution with short data acquisition. The training dataset is comprised of spatiotemporal ultrafast ultrasound signals acquired from in vivo mouse brains, while the testing dataset includes in vivo mouse brain, chicken embryo chorioallantoic membrane (CAM), and healthy human subjects. The in vivo mouse imaging studies demonstrate that CS-PD could achieve an approximate twofold improvement in spatial resolution when compared with conventional power Doppler. In addition, the microvascular images generated by CS-PD showed good agreement with the corresponding ULM images as indicated by a structural similarity index of 0.7837 and a peak signal-to-noise ratio (PSNR) of 25.52. Moreover, CS-PD was able to preserve the temporal profile of the blood flow (e.g., pulsatility) that is similar to conventional power Doppler. Finally, the generalizability of CS-PD was demonstrated on testing data of different tissues using different imaging settings. The fast inference time of the proposed deep neural network also allows CS-PD to be implemented for real-time imaging. These features of CS-PD offer a practical, fast, and robust microvascular imaging solution for many preclinical and clinical applications of Doppler ultrasound.

Deep learning based sarcopenia prediction from shear-wave ultrasonographic elastography and gray scale ultrasonography of rectus femoris muscle.

We aim to evaluate the performance of a deep convolutional neural network (DCNN) in predicting the presence or absence of sarcopenia using shear-wave elastography (SWE) and gray-scale ultrasonography (GSU) of rectus femoris muscle as an imaging biomarker. This retrospective study included 160 pair sets of GSU and SWE images (n = 160) from December 2018 and July 2019. Two radiologists scored the echogenicity of muscle on GSU (4-point score). Among them, 141 patients underwent CT and their L3 skeletal muscle index (SMI) were measured to categorize the presence or absence of sarcopenia. For DCNN, we used three CNN architectures (VGG19, ResNet-50, DenseNet 121). The accuracies of DCNNs for sarcopenia classification were 70.0-80.0% (based on SWE) and 65.0-75.0% (based on GSU). The DCNN application to SWE images highlights the utility of deep-learning base SWE for sarcopenia prediction. DCNN application to SWE images might be a potentially useful biomarker to predict sarcopenic status.

Arterial enhancing local tumor progression detection on CT images using convolutional neural network after hepatocellular carcinoma ablation: a preliminary study.

To evaluate the performance of a deep convolutional neural network (DCNN) in detecting local tumor progression (LTP) after tumor ablation for hepatocellular carcinoma (HCC) on follow-up arterial phase CT images. The DCNN model utilizes three-dimensional (3D) patches extracted from three-channel CT imaging to detect LTP. We built a pipeline to automatically produce a bounding box localization of pathological regions using a 3D-CNN trained for classification. The performance metrics of the 3D-CNN prediction were analyzed in terms of accuracy, sensitivity, specificity, positive predictive value (PPV), area under the receiver operating characteristic curve (AUC), and average precision. We included 34 patients with 49 LTP lesions and randomly selected 40 patients without LTP. A total of 74 patients were randomly divided into three sets: training (n = 48; LTP: no LTP = 21:27), validation (n = 10; 5:5), and test (n = 16; 8:8). When used with the test set (160 LTP positive patches, 640 LTP negative patches), our proposed 3D-CNN classifier demonstrated an accuracy of 97.59%, sensitivity of 96.88%, specificity of 97.65%, and PPV of 91.18%. The AUC and precision-recall curves showed high average precision values of 0.992 and 0.96, respectively. LTP detection on follow-up CT images after tumor ablation for HCC using a DCNN demonstrated high accuracy and incorporated multichannel registration.

AI musculoskeletal clinical applications: how can AI increase my day-to-day efficiency?

Artificial intelligence (AI) is expected to bring greater efficiency in radiology by performing tasks that would otherwise require human intelligence, also at a much faster rate than human performance. In recent years, milestone deep learning models with unprecedented low error rates and high computational efficiency have shown remarkable performance for lesion detection, classification, and segmentation tasks. However, the growing field of AI has significant implications for radiology that are not limited to visual tasks. These are essential applications for optimizing imaging workflows and improving noninterpretive tasks. This article offers an overview of the recent literature on AI, focusing on the musculoskeletal imaging chain, including initial patient scheduling, optimized protocoling, magnetic resonance imaging reconstruction, image enhancement, medical image-to-image translation, and AI-aided image interpretation. The substantial developments of advanced algorithms, the emergence of massive quantities of medical data, and the interest of researchers and clinicians reveal the potential for the growing applications of AI to augment the day-to-day efficiency of musculoskeletal radiologists.

Deep Generative Adversarial Networks: Applications in Musculoskeletal Imaging.

In recent years, deep learning techniques have been applied in musculoskeletal radiology to increase the diagnostic potential of acquired images. Generative adversarial networks (GANs), which are deep neural networks that can generate or transform images, have the potential to aid in faster imaging by generating images with a high level of realism across multiple contrast and modalities from existing imaging protocols. This review introduces the key architectures of GANs as well as their technical background and challenges. Key research trends are highlighted, including: (a) reconstruction of high-resolution MRI; (b) image synthesis with different modalities and contrasts; (c) image enhancement that efficiently preserves high-frequency information suitable for human interpretation; (d) pixel-level segmentation with annotation sharing between domains; and (e) applications to different musculoskeletal anatomies. In addition, an overview is provided of the key issues wherein clinical applicability is challenging to capture with conventional performance metrics and expert evaluation. When clinically validated, GANs have the potential to improve musculoskeletal imaging. Keywords: Adults and Pediatrics, Computer Aided Diagnosis (CAD), Computer Applications-General (Informatics), Informatics, Skeletal-Appendicular, Skeletal-Axial, Soft Tissues/Skin © RSNA, 2021.

Artificial intelligence in musculoskeletal ultrasound imaging.

Ultrasonography (US) is noninvasive and offers real-time, low-cost, and portable imaging that facilitates the rapid and dynamic assessment of musculoskeletal components. Significant technological improvements have contributed to the increasing adoption of US for musculoskeletal assessments, as artificial intelligence (AI)-based computer-aided detection and computer-aided diagnosis are being utilized to improve the quality, efficiency, and cost of US imaging. This review provides an overview of classical machine learning techniques and modern deep learning approaches for musculoskeletal US, with a focus on the key categories of detection and diagnosis of musculoskeletal disorders, predictive analysis with classification and regression, and automated image segmentation. Moreover, we outline challenges and a range of opportunities for AI in musculoskeletal US practice.

Temporal Trends in Cervical Spine Curvature of South Korean Adults Assessed by Deep Learning System Segmentation, 2006-2018.

Importance: The loss of the physiologic cervical lordotic curve is a common degenerative disorder known to be associated with abnormal spinal alignment. However, the changing trends among sex and age groups has not yet been well established. Objective: To analyze the temporal trends in cervical curvature across sex and age groups using an automated deep learning system (DLS). Design, Setting, and Participants: A retrospective cross-sectional study was conducted using lateral cervical radiographs of 13 691 individuals from January 1, 2006, to December 31, 2018. The degree of anterior vertical curvature was approximated by a DLS approach and convexity measurement method. This population-based study used the Yonsei University College of Medicine Severance Hospital, Seoul, South Korea, cohort database to identify 13 691 consecutive adults (≥18 years of age) who underwent standing lateral radiography in inpatient and outpatient settings. Main Outcomes and Measures: The prevalence of kyphotic and straight cervical curve as well as the trends of degree of cervical curvature in 2006 to 2018 among sex and age groups were determined. The DLS performance was validated with quantitative metrics and compared with interobserver and intraobserver variations. Results: Automatic cervical spine segmentation was identified from lateral radiographs of 13 691 individuals (mean [SD] age, 49.9 [15.3] years; 8051 women [58.8%]). From 2006 to 2018, the decrease in the lordotic curve was significant across both sexes and age groups younger than 70 years, with the decrease more pronounced in women and successively younger generations (female, -0.05; 95% CI, -0.06 to -0.04; 18-29 years of age, -0.06; 95% CI, -0.08 to -0.04; 30-39 years of age, -0.06; 95% CI, -0.08 to -0.04; and 40-49 years of age, -0.05; 95% CI, -0.06 to -0.03; all P < .001). The prevalence of straight and kyphotic curvature had a significant increasing trend for both sexes and young generations, in which individuals 18 to 29 years of age generally had the highest prevalence rates during the study cycle (in 2018, kyphosis, 16.7%; 95% CI, 10.8%-22.5%; straight, 45.5%; 95% CI, 37.7%-53.3%). Similar trends were observed with longitudinal analysis of repeated measures of individuals, with more pronounced decreases in lordotic curvature observed among women and young adults. Conclusions and Relevance: This study suggests a significant, increasing loss of normal cervical lordotic curvature for both sexes and young adults that is greater in progressively younger cohorts and women. Further research is necessary to evaluate associations between neck pain and loss of cervical curvature and address the need for active promotion and practical interventions aimed at neck posture correction.