ABSTRACT
Background: Persistent symptoms affect a subset of coronavirus disease 2019 (COVID-19) survivors. Some of these may be cardiovascular (CV)-related. Objective: To assess the burden of objective CV morbidity among, and to explore the short-term course experienced by, COVID-19 patients with post-infectious symptomatology suspected as CV. Methods: This was a single-center, retrospective analysis of consecutive adult patients with new-onset symptoms believed to be CV following recovery from COVID-19, who had been assessed at a dedicated 'Cardio'-COVID clinic between June 2020 and June 2021. All participants were followed for 1 year for symptomatic course and the occurrence of new CV diagnoses and major adverse cardiovascular events (MACE). Results: A total of 96 patients (median age 54 (IQR, 44-64) years, 52 (54%) females) were included in the final analysis. Initial visits occurred within a median of 142 days after the diagnosis of acute COVID. Nearly all (99%) patients experienced a symptomatic acute illness, which was graded as severe in 26 (27%) cases according to the National Institutes of Health (NIH) criteria. Long-COVID symptoms included mainly dyspnea and fatigue. While the initial work-up was mostly normal, 45% of the 11 cardiac magnetic resonance studies performed revealed pathologies. New CV diagnoses were made in nine (9%) patients and mainly included myocarditis that later resolved. An abnormal spirometry was the only variable associated with these. No MACE were recorded. Fifty-two (54%) participants felt that their symptoms improved. No association was found between CV morbidity and symptomatic course. Conclusions: In our experience, long-COVID symptoms of presumed CV origin signified actual CV disease in a minority of patients who, irrespective of the final diagnosis, faced a fair 1-year prognosis.
ABSTRACT
Lung ultrasound (LUS) is a cheap, safe and non-invasive imaging modality that can be performed at patient bed-side. However, to date LUS is not widely adopted due to lack of trained personnel required for interpreting the acquired LUS frames. In this work we propose a framework for training deep artificial neural networks for interpreting LUS, which may promote broader use of LUS. When using LUS to evaluate a patient's condition, both anatomical phenomena (e.g., the pleural line, presence of consolidations), as well as sonographic artifacts (such as A- and B-lines) are of importance. In our framework, we integrate domain knowledge into deep neural networks by inputting anatomical features and LUS artifacts in the form of additional channels containing pleural and vertical artifacts masks along with the raw LUS frames. By explicitly supplying this domain knowledge, standard off-the-shelf neural networks can be rapidly and efficiently finetuned to accomplish various tasks on LUS data, such as frame classification or semantic segmentation. Our framework allows for a unified treatment of LUS frames captured by either convex or linear probes. We evaluated our proposed framework on the task of COVID-19 severity assessment using the ICLUS dataset. In particular, we finetuned simple image classification models to predict per-frame COVID-19 severity score. We also trained a semantic segmentation model to predict per-pixel COVID-19 severity annotations. Using the combined raw LUS frames and the detected lines for both tasks, our off-the-shelf models performed better than complicated models specifically designed for these tasks, exemplifying the efficacy of our framework.