Dataset Information

Quantitative chest CT combined with plasma cytokines predict outcomes in COVID-19 patients.

ABSTRACT: Despite extraordinary international efforts to dampen the spread and understand the mechanisms behind SARS-CoV-2 infections, accessible predictive biomarkers directly applicable in the clinic are yet to be discovered. Recent studies have revealed that diverse types of assays bear limited predictive power for COVID-19 outcomes. Here, we harness the predictive power of chest CT in combination with plasma cytokines using a machine learning approach for predicting death during hospitalization and maximum severity degree in COVID-19 patients. Patients (n=152) from the Mount Sinai Health System in New York with plasma cytokine assessment and a chest CT within 5 days from admission were included. Demographics, clinical, and laboratory variables, including plasma cytokines (IL-6, IL-8, and TNF-α) were collected from the electronic medical record. We found that chest CT combined with plasma cytokines were good predictors of death (AUC 0.78) and maximum severity (AUC 0.82), whereas CT quantitative was better at predicting severity (AUC 0.81 vs 0.70) while cytokine measurements better predicted death (AUC 0.70 vs 0.66). Finally, we provide a simple scoring system using plasma IL-6, IL-8, TNF-α, GGO to aerated lung ratio and age as novel metrics that may be used to monitor patients upon hospitalization and help physicians make critical decisions and considerations for patients at high risk of death for COVID-19.

SUBMITTER: Carbonell G

PROVIDER: S-EPMC8528085 | biostudies-literature | 2021 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Quantitative chest CT combined with plasma cytokines predict outcomes in COVID-19 patients.

Carbonell Guillermo G Del Valle Diane Marie DM Gonzalez-Kozlova Edgar E Marinelli Brett B Klein Emma E El Homsi Maria M Stocker Daniel D Chung Michael M Bernheim Adam A Simons Nicole W NW Xiang Jiani J Nirenberg Sharon S Kovatch Patricia P Lewis Sara S Merad Miriam M Gnjatic Sacha S Taouli Bachir B

medRxiv : the preprint server for health sciences 20211014

Despite extraordinary international efforts to dampen the spread and understand the mechanisms behind SARS-CoV-2 infections, accessible predictive biomarkers directly applicable in the clinic are yet to be discovered. Recent studies have revealed that diverse types of assays bear limited predictive power for COVID-19 outcomes. Here, we harness the predictive power of chest CT in combination with plasma cytokines using a machine learning approach for predicting death during hospitalization and ma ...[more]

PMID: 34671777

Similar Datasets

Project description:INTRODUCTION:COVID-19 pneumonia is characterized by ground-glass opacities (GGOs) and consolidations on Chest CT, although these CT features cannot be considered specific, at least on a qualitative analysis. The aim is to evaluate if Quantitative Chest CT could provide reliable information in discriminating COVID-19 from non-COVID-19 patients. MATERIALS AND METHODS:From March 31, 2020 until April 18, 2020, patients with Chest CT suggestive for interstitial pneumonia were retrospectively enrolled and divided into two groups based on positive/negative COVID-19 RT-PCR results. Patients with pulmonary resection and/or CT motion artifacts were excluded. Quantitative Chest CT analysis was performed with a dedicated software that provides total lung volume, healthy parenchyma, GGOs, consolidations and fibrotic alterations, expressed both in liters and percentage. Two radiologists in consensus revised software analysis and adjusted areas of lung impairment in case of non-adequate segmentation. Data obtained were compared between COVID-19 and non-COVID-19 patients and p?<?0.05 were considered statistically significant. Performance of statistically significant parameters was tested by ROC curve analysis. RESULTS:Final population enrolled included 190 patients: 136 COVID-19 patients (87 male, 49 female, mean age 66?±?16) and 54 non-COVID-19 patients (25 male, 29 female, mean age 63?±?15). Lung quantification in liters showed significant differences between COVID-19 and non-COVID-19 patients for GGOs (0.55?±?0.26L vs 0.43?±?0.23L, p?=?0.0005) and fibrotic alterations (0.05?±?0.03 L vs 0.04?±?0.03 L, p?<?0.0001). ROC analysis of GGOs and fibrotic alterations showed an area under the curve of 0.661 (cutoff 0.39 L, 68% sensitivity and 59% specificity, p?<?0.001) and 0.698 (cutoff 0.02 L, 86% sensitivity and 44% specificity, p?<?0.001), respectively. CONCLUSIONS:Quantification of GGOs and fibrotic alterations on Chest CT could be able to identify patients with COVID-19.

Project description:ObjectivesTo develop and validate a machine learning model for the prediction of adverse outcomes in hospitalized patients with COVID-19.MethodsWe included 424 patients with non-severe COVID-19 on admission from January 17, 2020, to February 17, 2020, in the primary cohort of this retrospective multicenter study. The extent of lung involvement was quantified on chest CT images by a deep learning-based framework. The composite endpoint was the occurrence of severe or critical COVID-19 or death during hospitalization. The optimal machine learning classifier and feature subset were selected for model construction. The performance was further tested in an external validation cohort consisting of 98 patients.ResultsThere was no significant difference in the prevalence of adverse outcomes (8.7% vs. 8.2%, p = 0.858) between the primary and validation cohorts. The machine learning method extreme gradient boosting (XGBoost) and optimal feature subset including lactic dehydrogenase (LDH), presence of comorbidity, CT lesion ratio (lesion%), and hypersensitive cardiac troponin I (hs-cTnI) were selected for model construction. The XGBoost classifier based on the optimal feature subset performed well for the prediction of developing adverse outcomes in the primary and validation cohorts, with AUCs of 0.959 (95% confidence interval [CI]: 0.936-0.976) and 0.953 (95% CI: 0.891-0.986), respectively. Furthermore, the XGBoost classifier also showed clinical usefulness.ConclusionsWe presented a machine learning model that could be effectively used as a predictor of adverse outcomes in hospitalized patients with COVID-19, opening up the possibility for patient stratification and treatment allocation.Key points• Developing an individually prognostic model for COVID-19 has the potential to allow efficient allocation of medical resources. • We proposed a deep learning-based framework for accurate lung involvement quantification on chest CT images. • Machine learning based on clinical and CT variables can facilitate the prediction of adverse outcomes of COVID-19.

Project description:Background The standard for diagnosis of SARS-CoV-2 virus is reverse transcription polymerase chain reaction (RT-PCR) test, but chest CT may play a complimentary role in the early detection of COVID-19 pneumonia. Purpose To investigate CT features of patients with COVID-19 in Rome, Italy, and to compare the accuracy of CT with RT-PCR. Methods In this prospective study from March 4, 2020, until March 19, 2020, consecutive patients with suspected COVID-19 infection and respiratory symptoms were enrolled. Exclusion criteria were: chest CT with contrast medium performed for vascular indications, patients who refused chest CT or hospitalization, and severe CT motion artifact. All patients underwent RT-PCR and chest CT. Diagnostic performance of CT was calculated using RT-PCR as reference. Chest CT features were calculated in a subgroup of RT-PCR-positive and CT-positive patients. CT features of hospitalized patients and patient in home isolation were compared by using Pearson chi squared test. Results Our study population comprised 158 consecutive study participants (83 male and 75 female, mean age 57 y ±17). Fever was observed in 97/158 (61%), cough in 88/158 (56%), dyspnea in 52/158 (33%), lymphocytopenia in 95/158 (60%), increased C-reactive protein level in 139/158 (88%), and elevated lactate dehydrogenase in 128/158 (81%) study participants. Sensitivity, specificity, and accuracy of CT were 97% (60/62)[95% IC, 88-99%], 56% (54/96)[95% IC,45-66%] and 72% (114/158)[95% IC 64-78%], respectively. In the subgroup of RT-PCR-positive and CT-positive patients, ground-glass opacities (GGO) were present in 58/58 (100%), multilobe and posterior involvement were both present in 54/58 (93%), bilateral pneumonia in 53/58 (91%), and subsegmental vessel enlargement (> 3 mm) in 52/58 (89%) of study participants. Conclusion The typical pattern of COVID-19 pneumonia in Rome, Italy, was peripherally ground-glass opacities with multilobe and posterior involvement, bilateral distribution, and subsegmental vessel enlargement (> 3 mm). Chest CT sensitivity was high (97%) but with lower specificity (56%).

Dataset Information

Quantitative chest CT combined with plasma cytokines predict outcomes in COVID-19 patients.

Publications

Quantitative chest CT combined with plasma cytokines predict outcomes in COVID-19 patients.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets