Artificial intelligence guided HRCT assessment predicts the severity of COVID-19 pneumonia based on clinical parameters.

BACKGROUND: The purpose of the study was to compare the results of AI (artificial intelligence) analysis of the extent of pulmonary lesions on HRCT (high resolution computed tomography) images in COVID-19 pneumonia, with clinical data including laboratory markers of inflammation, to verify whether AI HRCT assessment can predict the clinical severity of COVID-19 pneumonia. METHODS: The analyzed group consisted of 388 patients with COVID-19 pneumonia, with automatically analyzed HRCT parameters of volume: AIV (absolute inflammation), AGV (absolute ground glass), ACV (absolute consolidation), PIV (percentage inflammation), PGV (percentage ground glass), PCV (percentage consolidation). Clinical data included: age, sex, admission parameters: respiratory rate, oxygen saturation, CRP (C-reactive protein), IL6 (interleukin 6), IG - immature granulocytes, WBC (white blood count), neutrophil count, lymphocyte count, serum ferritin, LDH (lactate dehydrogenase), NIH (National Institute of Health) severity score; parameters of clinical course: in-hospital death, transfer to the ICU (intensive care unit), length of hospital stay. RESULTS: The highest correlation coefficients were found for PGV, PIV, with LDH (respectively 0.65, 0.64); PIV, PGV, with oxygen saturation (respectively - 0.53, -0.52); AIV, AGV, with CRP (respectively 0.48, 0.46); AGV, AIV, with ferritin (respectively 0.46, 0.45). Patients with critical pneumonia had significantly lower oxygen saturation, and higher levels of immune-inflammatory biomarkers on admission. The radiological parameters of lung involvement proved to be strong predictors of transfer to the ICU (in particular, PGV ≥ cut-off point 29% with Odds Ratio (OR): 7.53) and in-hospital death (in particular: AIV ≥ cut-off point 831 cm3 with OR: 4.31). CONCLUSIONS: Automatic analysis of HRCT images by AI may be a valuable method for predicting the severity of COVID-19 pneumonia. The radiological parameters of lung involvement correlate with laboratory markers of inflammation, and are strong predictors of transfer to the ICU and in-hospital death from COVID-19. TRIAL REGISTRATION: National Center for Research and Development CRACoV-HHS project, contract number SZPITALE-JEDNOIMIENNE/18/2020.

The value of lung ultrasound in COVID-19 pneumonia, verified by high resolution computed tomography assessed by artificial intelligence.

BACKGROUND: Lung ultrasound (LUS) is an increasingly popular imaging method in clinical practice. It became particularly important during the COVID-19 pandemic due to its mobility and ease of use compared to high-resolution computed tomography (HRCT). The objective of this study was to assess the value of LUS in quantifying the degree of lung involvement and in discrimination of lesion types in the course of COVID-19 pneumonia as compared to HRCT analyzed by the artificial intelligence (AI). METHODS: This was a prospective observational study including adult patients hospitalized due to COVID-19 in whom initial HRCT and LUS were performed with an interval < 72 h. HRCT assessment was performed automatically by AI. We evaluated the correlations between the inflammation volume assessed both in LUS and HRCT, between LUS results and the HRCT structure of inflammation, and between LUS and the laboratory markers of inflammation. Additionally we compared the LUS results in subgroups depending on the respiratory failure throughout the hospitalization. RESULTS: Study group comprised 65 patients, median 63 years old. For both lungs, the median LUS score was 19 (IQR-interquartile range 11-24) and the median CT score was 22 (IQR 16-26). Strong correlations were found between LUS and CT scores (for both lungs r = 0.75), and between LUS score and percentage inflammation volume (PIV) (r = 0.69). The correlations remained significant, if weakened, for individual lung lobes. The correlations between LUS score and the value of the percentage consolidation volume (PCV) divided by percentage ground glass volume (PGV), were weak or not significant. We found significant correlation between LUS score and C-reactive protein (r = 0.55), and between LUS score and interleukin 6 (r = 0.39). LUS score was significantly higher in subgroups with more severe respiratory failure. CONCLUSIONS: LUS can be regarded as an accurate method to evaluate the extent of COVID-19 pneumonia and as a promising tool to estimate its clinical severity. Evaluation of LUS in the assessment of the structure of inflammation, requires further studies in the course of the disease. TRIAL REGISTRATION: The study has been preregistered 13 Aug 2020 on clinicaltrials.gov with the number NCT04513210.

The role of artificial intelligence technology analysis of HRCT images in predicting the severity of COVID-19 pneumonia.

INTRODUCTION: The purpose of the study was to analyze the role of automatic assessment of COVID-19 pneumonia severity in high resolution computed tomography (HRCT) images by artificial intelligence (AI) technology. PATIENTS AND METHODS: We retrospectively studied the medical records of consecutive patients admitted to the Krakow University Hospital due to COVID-19. Of the 1,729 patients, 804 had HRCT with automatically analyzed radiological parameters: absolute inflammation volume (AIV), absolute ground glass volume (AGV), absolute consolidation volume (ACV), percentage inflammation volume (PIV), percentage ground glass volume (PGV), percentage consolidation volume (PCV) and severity of pneumonia classified as none, mild, moderate, or critical. RESULTS: The automatically assessed radiological parameters correlated with the clinical parameters that reflected the severity of pneumonia (p < 0.05). Patients with critical pneumonia, compared to mild or moderate, were more frequently men, had significantly lower oxygen saturation, higher respiratory rate, higher levels of inflammatory markers, more common need for mechanical ventilation, and admission to the intensive care unit (ICU); moreover, they were more likely to die during hospitalization. Notably, as determined by the receiving operating characteristic curve, radiological parameters above or equal the cut-off points were independently associated with in-hospital mortality (ACV odds ratio (OR) 4.08, 95% confidence limits (CI) 2.62 - 6.35; PCV OR 4.05, CI 2.60 - 6.30). CONCLUSIONS: Using AI to analyze HRCT images is a simple and valuable approach to predict the severity of COVID-19 pneumonia.

Mechanical thrombectomy in COVID-19-associated ischaemic stroke: patient characteristics and outcomes in a single-centre study.

INTRODUCTION: The aim of this study was to assess the clinical profiles and outcomes of patients with confirmed COVID-19 infection and acute ischaemic stroke (AIS) treated with mechanical thrombectomy (MT) at the Comprehensive Stroke Centre (CSC) of the University Hospital in Krakow. CLINICAL RATIONALE FOR THE STUDY: COVID-19 is a risk factor for AIS and worsens prognosis in patients with large artery occlusions. During the pandemic, the global number of MT has dropped. At the same time, studies assessing outcomes of this treatment in COVID-19-associated AIS have produced divergent results. MATERIAL AND METHODS: In this single-centre study, we retrospectively analysed and compared the clinical profiles (age, sex, presence of cardiovascular risk factors, neurological deficit at admission), stroke size (measured using postprocessing analysis of perfusion CT with RAPID software), time from stroke onset to arrival at the CSC, time from arrival at the CSC to groin puncture, treatment with intravenous thrombolysis, length of hospitalisation, laboratory test results, and short-term outcomes (measured with Thrombolysis in Cerebral Infarction scale, modified Rankin Scale and National Health Institute Stroke Scale) in patients with AIS treated with MT during the pandemic. A comparison between patients with and without concomitant SARS-CoV2 infection was then performed. RESULTS: There were no statistically significant differences between 15 COVID (+) and 167 COVID (-) AIS patients treated with AIS with respect to clinical profiles (p > 0.05), stroke size (p > 0.05) or outcomes (NIHSS at discharge, 8.1 (SD = 7.1) vs. 8.8 (SD = 9.6), p = 0.778, mRS at discharge 2.9 (SD = 2) vs. 3.1 (SD = 2.1), p = 0.817, death rate 6.7% vs. 12.6%, p = 0.699). There was a significant difference between patients with and without COVID-19 concerning time from arrival at the CSC to groin puncture [104.27 (SD = 51.47) vs. 97.63 (SD = 156.94) min., p = 0.044] and the length of hospitalisation [23.7 (SD = 11.9) vs. 10.5 (SD = 6.9) days, p < 0.001]. CONCLUSION: In AIS patients treated with MT, concomitant SARS-CoV2 infection did not affect the outcome. Our observations need to be confirmed in larger, and preferably multicentre, studies.

CRACoV-HHS: an interdisciplinary project for multi-specialist hospital and non-hospital care for patients with SARS-CoV-2 infection as well hospital staff assessment for infection exposure.

The complex course of the COVID-19 and the distant complications of the SARS-CoV-2 infection still remain an unfaded challenge for modern medicine. The care of patients with the symptomatic course of COVID-19 exceeds the competence of a single specialty, often requiring a multispecialist approach. The CRACoV-HHS (CRAcow in CoVid pandemic - Home, Hospital and Staff) project has been developed by a team of scientists and clinicians with the aim of optimizing medical care at hospital and ambulatory settings and treatment of patients with SARS-CoV-2 infection. The CRACoV project integrates 26 basic and clinical research from multiple medical disciplines, involving different populations infected with SARS-CoV-2 virus and exposed to infection. Between January 2021 and April 2022 we plan to recruit subjects among patients diagnosed and treated in the University Hospital in Cracow, the largest public hospital in Poland, i.e. 1) patients admitted to the hospital due to COVID-19 [main module: 'Hospital']; 2) patients with signs of infection who have been confirmed as having SARS-CoV-2 infection and have been referred to home isolation due to their mild course (module: 'Home isolation'); 3) patients with symptoms of infection and high exposure to SARS- CoV-2 who have a negative RT-PCR test result. In addition, survey in various professional groups of hospital employees, both medical and non-medical, and final-fifth year medical students (module: 'Staff') is planned. The project carries both scientific and practical dimension and is expected to develop a multidisciplinary model of care of COVID-19 patients as well as recommendations for the management of particular groups of patients including: asymptomatic patient or with mild symptoms of COVID-19; symptomatic patients requiring hospitalization due to more severe clinical course of disease and organ complications; patient requiring surgery; patient with diabetes; patient requiring psychological support; patient with undesirable consequences of pharmacological treatment.

Differences among COVID-19, Bronchopneumonia and Atypical Pneumonia in Chest High Resolution Computed Tomography Assessed by Artificial Intelligence Technology.

The aim of this study was to compare the results of automatic assessment of high resolution computed tomography (HRCT) by artificial intelligence (AI) in 150 patients from three subgroups: pneumonia in the course of COVID-19, bronchopneumonia and atypical pneumonia. The volume percentage of inflammation and the volume percentage of "ground glass" were significantly higher in the atypical (respectively, 11.04%, 8.61%) and the COVID-19 (12.41%, 10.41%) subgroups compared to the bronchopneumonia (5.12%, 3.42%) subgroup. The volume percentage of consolidation was significantly higher in the COVID-19 (2.95%) subgroup compared to the atypical (1.26%) subgroup. The percentage of "ground glass" in the volume of inflammation was significantly higher in the atypical (89.85%) subgroup compared to the COVID-19 (79.06%) subgroup, which in turn was significantly higher compared to the bronchopneumonia (68.26%) subgroup. HRCT chest images, analyzed automatically by artificial intelligence software, taking into account the structure including "ground glass" and consolidation, significantly differ in three subgroups: COVID-19 pneumonia, bronchopneumonia and atypical pneumonia. However, the partial overlap, particularly between COVID-19 pneumonia and atypical pneumonia, may limit the usefulness of automatic analysis in differentiating the etiology. In our future research, we plan to use artificial intelligence for objective assessment of the dynamics of pulmonary lesions during COVID-19 pneumonia.

COVID-19 Infection Negative in Nasopharyngeal Swabs but Suspected in Computed Tomography and Confirmed in Bronchoalveolar Lavage Material.

We present a case of a patient with clinical symptoms of pneumonia, negative in several polymerase chain reaction COVID-19 tests from nasopharyngeal swabs but suspected in computed tomography and finally confirmed in bronchoalveolar lavage material.