Temporal radiographic changes in COVID-19 patients: relationship to disease severity and viral clearance

Liu, Xiaofan; Zhou, Hong; Zhou, Yilu; Wu, Xiaojun; Zhao, Yang; Lü, Yang; Tan, Weijun; Yuan, Mingli; Ding, Xuhong; Zou, Jinjing; Li, Ruiyun; Liu, Hailing; Ewing, Rob M.; Hu, Yi; Nie, Hanxiang; Wang, Yihua

doi:10.1038/s41598-020-66895-w

Cited by 35 publications

(37 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Similarly, in data in the USA (March 31, 2020), chronic respiratory diseases were comorbidities in 8.5% of patients with COVID-19, compared to the GBD estimate of 11.3% for the same disease 5 . There are also several published studies showing the synergistic effect of CLD in worsening the severity of COVID-19 [6][7][8][9][10] . Guan et al reported more than 50% of chronic pulmonary disease for COVID-19 patients who were also admitted to ICU 11 .…”

Section: Comorbiditiesmentioning

confidence: 99%

County-Level Longitudinal Clustering of COVID-19 Mortality to Incidence Ratio in the United States

Vahabi¹,

Salehi²,

Duarte³

et al. 2020

Preprint

View full text Add to dashboard Cite

Abstract Background: As of June 3, 2020, the mortality to incidence ratio (MIR) of COVID-19 was 5.8%. We utilized a longitudinal model-based clustering system based on the disease trajectories over time.Methods: County-level COVID-19 cases and deaths (March-June 2020), and a set of potential risk factors were collected for 3050 U.S. counties. We used growth mixture models to identify clusters of counties exhibiting similar COVID-19 MIR growth trajectories and risk-factors over time. Results: We identified clusters 1 (rural-areas of IA, NC, OK, VA, FL, GA, LA, OH states), and 7 (rural-areas of AR, CO, GA, KS, NE, TN, TX states) as the so-called “more vulnerable” clusters. Tuberculosis (OR=1.3), drug use disorder (OR=1.1), and particulate matter (OR=1.1) were significantly associated with increased odds of being in a more vulnerable cluster. Heart complications and cancer were the main risk factors increasing the COVID-19 MIR (range: 0.08%-0.72% MIR↑).Conclusion: We identified two county-clusters exhibiting the highest COVID-19 MIR trajectories, indicating that enhancing the capacity and access to healthcare resources would be key to successfully manage COVID-19 in these “vulnerable” clusters. These findings provide insights for public health policymakers on the groups of people and locations they need to pay particular attention while managing the COVID-19 epidemic.

show abstract

Section: Comorbiditiesmentioning

confidence: 99%

County-Level Longitudinal Clustering of COVID-19 Mortality to Incidence Ratio in the United States

Vahabi¹,

Salehi²,

Duarte³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Similarly, in data in the USA (Mar 31, 2020), chronic respiratory diseases were comorbidities in 8.5% of patients with COVID-19, compare to the GBD estimate of 11.3% for the same disease [5] . Several published studies show the synergistic effect of CLD in worsening the severity of COVID-19 [6][7][8][9][10] . Guan et al reported that more than 50% of chronic pulmonary disease for COVID -19 patients admitted to the ICU [11] .…”

Section: Comorbiditiesmentioning

confidence: 99%

County-Level Longitudinal Clustering of COVID-19 Mortality to Incidence Ratio in the United States

Vahabi¹,

Salehi²,

Duarte³

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: As of November 12, 2020, the mortality to incidence ratio (MIR) of COVID-19 was 5.8% in the US. We utilized a longitudinal model-based clustering system based on the disease trajectories over time. We aimed to find the so-called “vulnerable” cluster of counties where to dedicate additional resources by the US policymakers. Methods: County-level COVID-19 cases and deaths (Mar-Nov 2020), and a set of potential risk factors were collected for 3050 U.S. counties during the 1st wave (Mar25-Jun3, 2020), 1344 counties (sunbelt region) during the 2nd wave (Jun4-Sep2, 2020), and 1055 counties (great plains) during the 3rd wave (Sep3-Nov12, 2020). We used growth mixture models to identify clusters of counties exhibiting similar COVID-19 MIR growth trajectories and risk-factors over time. Results: We identified the so-called “more vulnerable” clusters during the 1st, 2nd and 3rd waves of COVID-19. Tuberculosis (OR=1.3-2.1-3.2), drug use disorder (OR=1.1), hepatitis (OR=13.1), HIV/AIDS (OR=2.3), cardiomyopathy and myocarditis (OR=1.3), diabetes (OR=1.2), mesothelioma (OR=9.3) were significantly associated with increased odds of being in a more vulnerable cluster. Heart complications and cancer were the main risk factors increasing the COVID-19 MIR (range: 0.08%-0.52% MIR↑).Conclusion: We identified the so-called “more vulnerable” county-clusters exhibiting the highest COVID-19 MIR trajectories, indicating that enhancing the capacity and access to healthcare resources would be key to successfully manage COVID-19 in these clusters. These findings provide insights for public health policymakers on the groups of people and locations they need to pay particular attention while managing the COVID-19 epidemic.

show abstract

“…The morphological changes in the peripheral lungs of patients with moderate or severe disease [17][18][19][20] create imaging markers in chest radiographs. Chest imaging can play an important role in the clinical management of patients with COVID-19 that have clinical manifestations or worsening respiratory symptoms due to COVID-19 21,22 .…”

Section: Introductionmentioning

confidence: 99%

Clinical deployment and validation of a radiology artificial intelligence system for COVID-19

Hashemian

Manchanda

et al. 2020

Preprint

View full text Add to dashboard Cite

The global COVID-19 pandemic has disrupted patient care delivery in healthcare systems world-wide. For healthcare providers to better allocate their resources and improve the care for patients with severe disease, it is valuable to be able to identify those patients with COVID-19 who are at higher risk for clinical complications. This may help to optimize clinical workflow and more efficiently allocate scarce medical resources. To this end, medical imaging shows great potential and artificial intelligence (AI) algorithms have been developed to assist in diagnosing and risk stratifying COVID-19 patients. However, despite the rapid development of numerous AI models, these models cannot be clinically useful unless they can be deployed in real-world environments in real-time on clinical data. Here, we propose an end-to-end AI hospital-deployment architecture for COVID-19 medical imaging algorithms in hospitals. We have successfully implemented this system at our institution and it has been used in prospective clinical validation of a deep learning algorithm potentially useful for triaging of patients with COVID-19. We demonstrate that many orchestration processes are required before AI inference can be performed on a radiology studies in real-time with the AI model being just one of the components that make up the AI deployment system. We also highlight that failure of any one of these processes can adversely affect the model's performance.

show abstract

Temporal radiographic changes in COVID-19 patients: relationship to disease severity and viral clearance

Cited by 35 publications

References 25 publications

County-Level Longitudinal Clustering of COVID-19 Mortality to Incidence Ratio in the United States

County-Level Longitudinal Clustering of COVID-19 Mortality to Incidence Ratio in the United States

County-Level Longitudinal Clustering of COVID-19 Mortality to Incidence Ratio in the United States

Clinical deployment and validation of a radiology artificial intelligence system for COVID-19

Contact Info

Product

Resources

About