Abstract Our automated deep learning-based approach identifies consolidation/collapse in LUS images to aid in the diagnosisof 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 ona 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 ourprevious 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 expertsand improves over the video-based method of our previous work on pleural effusions.
Abstract The incidence and morbidity of deep venous thrombosis (DVT) and pulmonary embolus are high. Although efforts to increase screening for DVT have been recommended, this is limited by resources. Venous duplex ultrasound has replaced venography as the first‐line investigation of choice for DVT, increasing availability and reducing patient exposure to radiation and intravenous contrast. Furthermore, an abbreviated ultrasound where DVT is inferred from incomplete venous compressibility has an equivalent accuracy to venous duplex, requiring less time and training enabling its widespread use by emergency, critical care and anaesthesia clinicians. In this review, the evolution and method of lower limb venous compression ultrasound is described along with evidence for its use in patients at high risk for DVT in these clinical settings.
Background: Between 1964 and 1996, the 10-year survival of patients having valve replacement surgery for rheumatic heart disease (RHD) in the Northern Territory, Australia, was 68%. As medical care has evolved since then, this study aimed to determine whether there has been a corresponding improvement in the region’s RHD patients’ 10-year survival.Methods: A retrospective study of Aboriginal RHD patients in the Northern Territory, Australia, having their first valve surgery between 1997 and 2016. Survival was examined using Kaplan—Meier and Cox regression analysis.Findings: The cohort included 281 adults and 61 children. The median (interquartile range (IQR)) age at first surgery was 31 (18-42) years; 173/342 (51%) had a valve replacement. Of 281 adults, 204 (73%) had at least one preoperative comorbidity. There were 93/342 (27%) deaths during a median (IQR) follow-up of 8 (4-12) years. The overall 10-year survival was 70% [95% confidence interval (CI): 64-76], while it was 62% [95%CI (53-70)] in those having a valve replacement. Preoperative comorbidity was associated with earlier death, the risk of death increasing with each comorbidity (hazard ratio (HR): 1·3 (95%CI: 1·2-1·5), p<0·001). Preoperative chronic kidney disease (HR: 6·1 (95%CI: 2·8-13·3), p<0·001), coronary artery disease (HR: 2·3 (95%CI: 1·1-5·0), p=0·037) and pulmonary artery systolic pressure >50mmHg before surgery (HR: 2·1 (95%CI: 1·3-3·5), p=0·002) independently predicted death.Interpretation: Survival after RHD surgery in this region of Australia has not improved. Although the patients were young, many had multiple comorbidities which influenced long-term outcomes. The increasing prevalence of complex comorbidity in the region is a barrier to achieving optimal health outcomes.Funding Information: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Declaration of Interests: All co-authors declared no conflict of interest.Ethical Approval: The Human Research Ethics Committees of the Northern Territory Department of Health and Menzies School of Health Research (HREC reference number: 2016-25) and the Australian Institute of Health and Welfare (AIHW reference number: EO2018/5/410) provided ethical approval for the study. The committees waived the requirement for informed consent as the data were retrospective, de-identified at the time of analysis and presented in an aggregated manner.
The adoption of point-of-care lung ultrasound for both suspected and confirmed COVID-19 patients highlights the issues of accessibility to ultrasound training and equipment. Lung ultrasound is more sensitive than chest radiography in detecting viral pneumonitis and preferred over computed tomography for reasons including its portability, reduced healthcare worker exposure and repeatability. The main lung ultrasound findings in COVID-19 patients are interstitial syndrome, irregular pleural line and subpleural consolidations. Consolidations are most likely found in critical patients in need of ventilatory support. Hence, lung ultrasound may be used to timely triage patients who may have evolving pneumonitis. Other respiratory pathology that may be detected by lung ultrasound includes pulmonary oedema, pneumothorax, consolidation and large effusion. A key barrier to incorporate lung ultrasound in the assessment of COVID-19 patients is adequate decontamination of ultrasound equipment to avoid viral spread. This tutorial provides a practical method to learn lung ultrasound and a cost-effective method of preventing contamination of ultrasound equipment and a practical method for performing and interpreting lung ultrasound.
'/%3e%3c/defs%3e%3cpath fill='%23012169' d='M0 0h10080v5040H0z'/%3e%3cpath stroke='%23fff' stroke-width='.6' d='m0 0 6 3m0-3L0 3' clip-path='url(%23b)' transform='scale(840)'/%3e%3cpath stroke='%23e4002b' stroke-width='.4' d='m0 0 6 3m0-3L0 3' clip-path='url(%23c)' transform='scale(840)'/%3e%3cpath stroke='%23fff' stroke-width='840' d='M2520 0v2520M0 1260h5040'/%3e%3cpath stroke='%23e4002b' stroke-width='504' d='M2520 0v2520M0 1260h5040'/%3e%3cg fill='%23fff'%3e%3cuse xlink:href='%23d' x='2520' y='3780'/%3e%3cuse xlink:href='%23a' x='7560' y='4200'/%3e%3cuse xlink:href='%23a' x='6300' y='2205'/%3e%3cuse xlink:href='%23a' x='7560' y='840'/%3e%3cuse xlink:href='%23a' x='8680' y='1869'/%3e%3cuse xlink:href='%23e' x='8064' y='2730'/%3e%3c/g%3e%3c/svg%3e)
